Artificial Intelligence Software Sector Outlook — September 2026

Read the report

SlidesScroll

1 of 22

←→Full screen

INFORMATION TECHNOLOGY › SOFTWARE AND SERVICES › ARTIFICIAL INTELLIGENCE SOFTWARE

Artificial Intelligence Software: Value Splits by Model

The report shows where public valuations and precedent transactions place the greatest weight across the sector.

September 2026 · Prepared by NeuraCap AI · Confidential

Market data as of 2026-09-28 · primary valuation basis EV / Revenue (CY2027E)

Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice

Sources & methodology

1

← Previous

1 / 22 · INFORMATION TECHNOLOGY › SOFTWARE AND SERVICES › ARTIFICIAL INTELLIGENCE SOFTWARE

Next →

What each page shows

The analyst’s walkthrough of the deck, page by page.

1. 01

   INFORMATION TECHNOLOGY › SOFTWARE AND SERVICES › ARTIFICIAL INTELLIGENCE SOFTWARE

   INFORMATION TECHNOLOGY › SOFTWARE AND SERVICES › ARTIFICIAL INTELLIGENCE SOFTWARE.

   Cover — Artificial Intelligence Software sector insights, September 2026. Prepared by NeuraCap AI; market data as of 2026-09-28; primary valuation basis EV / Revenue (CY2027E). The framing line carries the deck's finding with the figures stripped. 1

   Everything on this page

   INFORMATION TECHNOLOGY › SOFTWARE AND SERVICES › ARTIFICIAL INTELLIGENCE SOFTWARE Artificial Intelligence Software: Value Splits by Model The report shows where public valuations and precedent transactions place the greatest weight across the sector. September 2026 · Prepared by NeuraCap AI · Confidential Market data as of 2026-09-28 · primary valuation basis EV / Revenue (CY2027E) Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 1

2. 02
   CONTENTS

   What This Report Covers

   What This Report Covers.

   Contents — the report runs 5 numbered sections plus the appendix: the bottom line, the landscape, valuation & situations, precedent transactions, strategic implications. The Bottom Line comes first by design: a reader who stops after section 01 still leaves with the whole story. 2

   Everything on this page

   CONTENTS What This Report Covers 01 The Bottom Line The Sector Contains Distinct Business Models with Sharply Different Valuation Profiles 02 The Landscape Business Model Matters as Much as the Artificial Intelligence Software Label 03 Valuation & Situations The Premium End Holds a Wide Lead Even After Forecast Growth Is Credited 04 Precedent Transactions Precedent Transactions Put a Wide Range Around Buyer Appetite 05 Strategic Implications Production Proof and Software Economics Shape the Operating Agenda 06 Appendix Comparables Detail, Methodology, Sources and Assumptions Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 2

3. 03
   01 · THE BOTTOM LINE

   Artificial Intelligence Software Spans Four Models, with Premiums Across Infrastructure and Services

   Artificial Intelligence Software Spans Four Models, with Premiums Across Infrastructure and Services.

   Artificial Intelligence Software Spans Four Models, with Premiums Across Infrastructure and Services The full story on one page · figures on EV / Revenue (CY2027E), market data as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Primary valuation basis: EV / Revenue on CY2027E consensus (16 of 17 companies eligible; multiples failing the platform's plausibility gates are excluded, never plotted). Practitioners price this industry on EV / Revenue rather than EV / EBITDA; validated coverage supports the industry standard (16 of 17 companies), so this report follows it. Qualitative characterisations are NeuraCap views. 3

   Everything on this page

   01 · THE BOTTOM LINE Artificial Intelligence Software Spans Four Models, with Premiums Across Infrastructure and Services The full story on one page · figures on EV / Revenue (CY2027E), market data as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Primary valuation basis: EV / Revenue on CY2027E consensus (16 of 17 companies eligible; multiples failing the platform's plausibility gates are excluded, never plotted). Practitioners price this industry on EV / Revenue rather than EV / EBITDA; validated coverage supports the industry standard (16 of 17 companies), so this report follows it. Qualitative characterisations are NeuraCap views. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 3 1 Forward Revenue Values Separate Sharply The premium end sits at 8.2x EV / Revenue, alongside 0.7x at the discount end. A forward multiple already credits forecast growth, so the remaining spread signals different market views on durability. 2 Faster Growth Sits with Higher Valuation On the 7 faster-growing names, the middle of the range is 5.1x, compared with 3.0x across the 6 slower-growing names. The association supports a focus on durable expansion rather than headline growth alone. 3 Business Model Changes the Reference Point AI infrastructure and model-enablement platforms sit at 3.2x across 6 names, while AI-native enterprise applications sit at 1.4x across 8 names. Revenue mix, deployment depth and inference economics remain important distinctions within those groups. 4 Revenue Remains the Practical Valuation Lens Forward EV / Revenue is available for 16 of the 17 approved companies, while 9 of 17 have meaningful forward EBITDA. That coverage supports revenue as the primary comparison, with profitability used as an execution cross-check. 3.0x Sector median EV/Revenue CY2027E consensus · EV/Revenue is the lens because practitioners price this growth set on revenue and 5 of 17 names are… 8.2x Premium end EV/Revenue vs 0.7x at the discount end top quartile (n=4) against bottom quartile (n=4) on EV/Revenue — the spread the report explains 12 Transactions with disclosed terms 70 recorded in this tier · 3 told as case studies, the full list in the appendix

4. 04
   SECTION 02

   02

   02.

   Section 02 of 06 — Business Model Matters as Much as the Artificial Intelligence Software Label. Applications, infrastructure, services and adjacent media offer different revenue quality and operating exposure. Use this divider to reset the room before the section's pages; the progress dots show where the argument stands. 4

   Everything on this page

   SECTION 02 02 THE LANDSCAPE Business Model Matters as Much as the Artificial Intelligence Software Label Applications, infrastructure, services and adjacent media offer different revenue quality and operating exposure. 02 of 06 Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 4

5. 05
   02 · MARKET MAP

   Valuation Spans Several Artificial Intelligence Software Models Rather than a Single Archetype

   Valuation Spans Several Artificial Intelligence Software Models Rather than a Single Archetype.

   Valuation Spans Several Artificial Intelligence Software Models Rather than a Single Archetype 17 approved companies grouped by business segment · median EV / Revenue (CY2027E) per group · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Segment grouping follows the platform's classification; group medians on rated names only. "Why it matters" lines are NeuraCap views. 5

   Everything on this page

   02 · MARKET MAP Valuation Spans Several Artificial Intelligence Software Models Rather than a Single Archetype 17 approved companies grouped by business segment · median EV / Revenue (CY2027E) per group · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Segment grouping follows the platform's classification; group medians on rated names only. "Why it matters" lines are NeuraCap views. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 5 AI-NATIVE ENTERPRISE APPLICATIONS 8 cos median 1.4x STRC SOUN AI RZLV CRNC VERI DUOT SSTI Workflow ownership can support recurring revenue, but pilot-to-production conversion and retention determine durability. AI INFRASTRUCTURE AND MODEL-ENABLEMENT PLATFORMS 6 cos median 3.2x IOT S AIOT GRRR PDYN AISP These platforms sit close to deployment, security and model orchestration, making architecture and inference economics central. AI IMPLEMENTATION AND MANAGED TECHNOLOGY SERVICES 2 cos median 5.1x INOD BBAI This group can benefit from enterprise adoption while carrying greater exposure to services mix and delivery intensity. ADJACENT: AI-ENABLED GAMING AND INTERACTIVE MEDIA 1 cos no rated names GDC The adjacent model adds audience and content exposure that differs from enterprise software economics.

6. 06
   02 · LANDSCAPE

   Infrastructure Carries the Higher Scaled Reference, While Applications Cover the Broadest Ground

   Infrastructure Carries the Higher Scaled Reference, While Applications Cover the Broadest Ground.

   Infrastructure Carries the Higher Scaled Reference, While Applications Cover the Broadest Ground Segment view of the approved universe · EV / Revenue (CY2027E) medians on rated names · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. "What they do / why it matters" is a NeuraCap view. Full company-level detail in the appendix and companion tables. 6

   Everything on this page

   02 · LANDSCAPE Infrastructure Carries the Higher Scaled Reference, While Applications Cover the Broadest Ground Segment view of the approved universe · EV / Revenue (CY2027E) medians on rated names · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. "What they do / why it matters" is a NeuraCap view. Full company-level detail in the appendix and companion tables. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 6 Segment n Share of universe Median EV/Revenue Names to know What they do — and why it matters AI-native enterprise applications 8 47% 1.4x Strategy Inc (STRC) · SoundHound AI, Inc. (SOUN) · +6 more Applications span the range. The 8-name group carries a 1.4x EV / Revenue reference and includes workflow products with different levels of recurring revenue, production deployment and customer concentration. AI infrastructure and model-enablement platforms 6 35% 3.2x Samsara Inc. (IOT) · SentinelOne, Inc. (S) · +4 more Infrastructure holds the higher reference. Across 6 names, the group sits at 3.2x EV / Revenue. Model-agnostic architecture, security, deployment depth and efficient inference economics can strengthen the commercial case. AI implementation and managed technology services 2 12% 5.1x Innodata Inc. (INOD) · BigBear.ai Holdings, Inc. (BBAI) Services connect adoption to delivery. The 2-name group gives owners a reference for businesses that combine technical capability with implementation. Revenue mix and delivery intensity remain central to the quality of that profile. Adjacent: AI-enabled gaming and interactive media 1 6% — GD Culture Group Limited (GDC) Adjacent exposure needs separate treatment. The single company sits outside the main enterprise software models and has no forward EV / Revenue estimate, limiting direct comparison with the rest of the peer set.

7. 07
   SECTION 03

   03

   03.

   Section 03 of 06 — The Premium End Holds a Wide Lead Even After Forecast Growth Is Credited. Forward EV / Revenue separates businesses with different expectations for durability and execution. Use this divider to reset the room before the section's pages; the progress dots show where the argument stands. 7

   Everything on this page

   SECTION 03 03 VALUATION & SITUATIONS The Premium End Holds a Wide Lead Even After Forecast Growth Is Credited Forward EV / Revenue separates businesses with different expectations for durability and execution. 03 of 06 Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 7

8. 08
   03 · PUBLIC MARKET VALUATION

   The Premium End Holds Its Lead After Forecast Growth Is Already Credited

   The Premium End Holds Its Lead After Forecast Growth Is Already Credited.

   The Premium End Holds Its Lead After Forecast Growth Is Already Credited EV / Revenue (CY2027E) · all 16 rated companies, sorted descending · sector median 3.0x · EV/Revenue is the lens because practitioners price this growth set on revenue and 5 of 17 names are loss-making on forward EBITDA · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Primary valuation basis: EV / Revenue on CY2027E consensus (16 of 17 companies eligible; multiples failing the platform's plausibility gates are excluded, never plotted). Practitioners price this industry on EV / Revenue rather than EV / EBITDA; validated coverage supports the industry standard (16 of 17 companies), so this report follows it. Tier zones are NeuraCap groupings cut at the rated set's quartiles;

   Everything on this page

   03 · PUBLIC MARKET VALUATION The Premium End Holds Its Lead After Forecast Growth Is Already Credited EV / Revenue (CY2027E) · all 16 rated companies, sorted descending · sector median 3.0x · EV/Revenue is the lens because practitioners price this growth set on revenue and 5 of 17 names are loss-making on forward EBITDA · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Primary valuation basis: EV / Revenue on CY2027E consensus (16 of 17 companies eligible; multiples failing the platform's plausibility gates are excluded, never plotted). Practitioners price this industry on EV / Revenue rather than EV / EBITDA; validated coverage supports the industry standard (16 of 17 companies), so this report follows it. Tier zones are NeuraCap groupings cut at the rated set's quartiles; every multiple quoted on this page is a median on the same EV / Revenue (CY2027E) basis. Panel commentary is a NeuraCap view. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 8 PREMIUM · median 8.2x CORE · median 3.0x DISCOUNT · median 0.7x Sector median 3.0x WHAT SEPARATES THE TWO ENDS The spread remains wide. The premium end sits at 8.2x EV / Revenue, compared with 0.7x at the discount end. That gap is observed across businesses with different products and operating profiles. Forward values demand durability. Forward EV / Revenue already incorporates forecast expansion. A premium that remains after that credit points to market confidence in revenue quality, deployment depth or longer-term economics. One label hides several models. Both ends contain AI-native enterprise applications, while infrastructure and managed technology services also appear across the ranking. Segment alone does not settle the valuation outcome.

9. 09
   03 · VALUATION DRIVERS

   Profitability Separates the Two Ends: Names Above the 17% Margin Line Carry 4.5x Against 0.9x Below It

   Profitability Separates the Two Ends: Names Above the 17% Margin Line Carry 4.5x Against 0.9x Below It.

   Profitability Separates the Two Ends: Names Above the 17% Margin Line Carry 4.5x Against 0.9x Below It Median EV / Revenue (CY2027E) by revenue-growth cohort and by EBITDA-margin cohort (each split at its covered median) · rated names with estimates · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Cohort medians on rated names with the required estimates (faster n=7; slower n=6; higher-margin n=5; lower-margin n=4). Driver readings are NeuraCap views on the supplied data — association, not causation. 9

   Everything on this page

   03 · VALUATION DRIVERS Profitability Separates the Two Ends: Names Above the 17% Margin Line Carry 4.5x Against 0.9x Below It Median EV / Revenue (CY2027E) by revenue-growth cohort and by EBITDA-margin cohort (each split at its covered median) · rated names with estimates · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Cohort medians on rated names with the required estimates (faster n=7; slower n=6; higher-margin n=5; lower-margin n=4). Driver readings are NeuraCap views on the supplied data — association, not causation. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 9 EV/Revenue, median per cohort · growth split at 18% · EBITDA-margin split at 17% Growth Separates the Two Observed Valuation Groups Among the 13 names with growth estimates used in the split, the 7 names at or above 18% sit at 5.1x, while the 6 names below that level sit at 3.0x. The result shows association rather than causation. Production Conversion Gives Growth Greater Commercial Weight Expansion tied to production deployment, retention and usage-based revenue is more durable than pilot-heavy activity. Owners can test growth quality through contracted backlog, remaining performance obligations and customer expansion. Inference Economics Remain Part of the Value Case Revenue growth has greater strategic value when compute, data-processing and support costs preserve software economics. Model-agnostic architecture and disciplined workload costs can improve that operating profile.

10. 10
    03 · SITUATION MAP

    A Premium Position Combines Above-Range Growth with an Above-Range Multiple

    A Premium Position Combines Above-Range Growth with an Above-Range Multiple.

    A Premium Position Combines Above-Range Growth with an Above-Range Multiple Cut on EV / Revenue vs the sector median (3.0x) (rows) and revenue growth vs the covered median (18%) (columns) · 3 rated names without the second measure are not mapped · observations, not recommendations Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Situation boundaries are the cohort's own medians (Directional, NeuraCap view). This page characterises situations; it does not recommend buying or selling any security. 10

    Everything on this page

    03 · SITUATION MAP A Premium Position Combines Above-Range Growth with an Above-Range Multiple Cut on EV / Revenue vs the sector median (3.0x) (rows) and revenue growth vs the covered median (18%) (columns) · 3 rated names without the second measure are not mapped · observations, not recommendations Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Situation boundaries are the cohort's own medians (Directional, NeuraCap view). This page characterises situations; it does not recommend buying or selling any security. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 10 Premium Growth Leaders Above-median multiple · above-median revenue growth 5 names Samsara Inc. (IOT) · SentinelOne, Inc. (S) · SoundHound AI, Inc. (SOUN) · +2 more Among the 5 mapped names in this position, growth and valuation both sit above the peer-set reference. The operating challenge is to sustain production deployment, retention and software economics. Premium with Slower Growth Above-median multiple · below-median revenue growth 3 names Strategy Inc (STRC) · C3.ai, Inc. (AI) · BigBear.ai Holdings, Inc. (BBAI) Among the 3 mapped names in this position, valuation sits above the peer-set reference while growth sits below it. Durability, workflow ownership or strategic positioning may be carrying greater weight than current expansion. Growth Awaiting Recognition Below-median multiple · above-median revenue growth 2 names Veritone, Inc. (VERI) · Airship AI Holdings, Inc. (AISP) Among the 2 mapped names in this position, growth sits above the peer-set reference while valuation remains below it. The gap focuses attention on revenue quality, cost structure and confidence in forecast delivery. Execution Reset Below-median multiple · below-median revenue growth 3 names PowerFleet, Inc. (AIOT) · Cerence Inc. (CRNC) · SoundThinking, Inc. (SSTI) Among the 3 mapped names in this position, both growth and valuation sit below the peer-set reference. The strategic agenda centers on production conversion, revenue mix and operating discipline.

11. 11
    03 · THE AGENDA

    Revenue Mix, Pricing and Cost Structure Can Strengthen Strategic Positioning

    Revenue Mix, Pricing and Cost Structure Can Strengthen Strategic Positioning.

    Revenue Mix, Pricing and Cost Structure Can Strengthen Strategic Positioning NeuraCap view · framed as the questions an owner or acquirer should resolve · observations, not recommendations Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. All content on this page is Directional — NeuraCap advisory judgment grounded in the cohort data shown earlier. It does not constitute investment advice. 11

    Everything on this page

    03 · THE AGENDA Revenue Mix, Pricing and Cost Structure Can Strengthen Strategic Positioning NeuraCap view · framed as the questions an owner or acquirer should resolve · observations, not recommendations Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. All content on this page is Directional — NeuraCap advisory judgment grounded in the cohort data shown earlier. It does not constitute investment advice. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 11 Deepen Production Deployment Prioritize products and customer workflows that move from pilots into recurring, embedded use. Greater workflow ownership can improve retention and revenue durability. What changes the answer: The answer changes when production conversion, usage expansion and renewal evidence become durable. Shift Revenue Toward Repeatable Software Assess where services support adoption and where they dilute repeatability. Productization, pricing and delivery design can improve revenue quality without removing necessary implementation support. What changes the answer: The answer changes when recurring software expands faster than delivery-intensive revenue. Protect Inference Economics Align architecture, pricing and workload design with compute and data-processing costs. Model-agnostic choices can reduce dependence on a single foundation model or infrastructure provider. What changes the answer: The answer changes when workload growth improves unit economics rather than increasing cost exposure. Strengthen Data and Model Defensibility Build around proprietary data access, clear rights and deep enterprise integration. These features can support differentiation as foundation models become more widely available. What changes the answer: The answer changes when customer contracts, data provenance and model dependencies support durable deployment.

12. 12
    SECTION 04

    04

    04.

    Section 04 of 06 — Precedent Transactions Put a Wide Range Around Buyer Appetite. Disclosed pricing varies with the target, strategic fit and earnings profile. Use this divider to reset the room before the section's pages; the progress dots show where the argument stands. 12

    Everything on this page

    SECTION 04 04 PRECEDENT TRANSACTIONS Precedent Transactions Put a Wide Range Around Buyer Appetite Disclosed pricing varies with the target, strategic fit and earnings profile. 04 of 06 Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 12

13. 13
    04 · DEAL CASE STUDIES

    Precedent Transactions Show Buyers Using Different Valuation Anchors

    Precedent Transactions Show Buyers Using Different Valuation Anchors.

    Precedent Transactions Show Buyers Using Different Valuation Anchors 3 of 12 transactions with disclosed terms, told as case studies · multiples on LTM financials at announcement where disclosed · the complete list is in the appendix · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. 66 precedent record(s) carry data-quality flags (deal value unit unresolved; divestiture roles reassigned; duplicate precedent id); figures are shown as recorded in the filing. 58 recorded transactions with neither a disclosed value nor a multiple are omitted from the precedent list and the case studies (kept in the companion workbook). Deal multiples are LTM at announcement (Definitive, from filings); they are not directly comparable to the CY2027E public basis and no spread is claimed.

    Everything on this page

    04 · DEAL CASE STUDIES Precedent Transactions Show Buyers Using Different Valuation Anchors 3 of 12 transactions with disclosed terms, told as case studies · multiples on LTM financials at announcement where disclosed · the complete list is in the appendix · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. 66 precedent record(s) carry data-quality flags (deal value unit unresolved; divestiture roles reassigned; duplicate precedent id); figures are shown as recorded in the filing. 58 recorded transactions with neither a disclosed value nor a multiple are omitted from the precedent list and the case studies (kept in the companion workbook). Deal multiples are LTM at announcement (Definitive, from filings); they are not directly comparable to the CY2027E public basis and no spread is claimed. "Why the deal happened" is a NeuraCap read of the recorded evidence. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 13 Sep-2025 $71M Scilex Holding Company Scilex Holding Company acquired Datavault AI Inc. at a high revenue reference. EV / LTM revenue 15.9x EV / LTM EBITDA n/a WHY THE DEAL HAPPENED The pairing suggests strategic interest in Datavault AI Inc.'s capabilities. The completed transaction indicates conviction beyond a commercial partnership. HOW THE TARGET WAS VALUED The disclosed value was $71M, equal to 15.9x EV / Revenue. That pricing provides a high-end transaction benchmark within the disclosed set. Jun-2021 $400M WiMi Hologram Cloud Inc. WiMi Hologram Cloud Inc. placed substantial value on VIYI Algorithm Inc. EV / LTM revenue n/a EV / LTM EBITDA n/a WHY THE DEAL HAPPENED The pairing suggests a strategic fit between WiMi Hologram Cloud Inc. and VIYI Algorithm Inc. The announced transaction points to interest in adding adjacent AI capability at scale. HOW THE TARGET WAS VALUED The disclosed value was $400M. With no transaction multiple stated, the value serves as the principal benchmark. Nov-2023 $70M BigBear.ai Holdings, Inc. BigBear.ai Holdings, Inc. (BBAI) expanded through Pangiam Intermediate Holdings, LLC. EV / LTM revenue n/a EV / LTM EBITDA n/a WHY THE DEAL HAPPENED The transaction suggests BigBear.ai Holdings, Inc. (BBAI) saw strategic fit in Pangiam Intermediate Holdings, LLC. The completed acquisition provides evidence of capability-led expansion within Artificial Intelligence Software. HOW THE TARGET WAS VALUED The disclosed value was $70M. The value benchmarks the transaction without relying on an unstated revenue or earnings multiple.

14. 14
    SECTION 05

    05

    05.

    Section 05 of 06 — Production Proof and Software Economics Shape the Operating Agenda. Owners can strengthen positioning through revenue quality, deployment depth and cost discipline. Use this divider to reset the room before the section's pages; the progress dots show where the argument stands. 14

    Everything on this page

    SECTION 05 05 STRATEGIC IMPLICATIONS Production Proof and Software Economics Shape the Operating Agenda Owners can strengthen positioning through revenue quality, deployment depth and cost discipline. 05 of 06 Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 14

15. 15
    05 · STRATEGIC IMPLICATIONS

    Owners Can Improve Positioning Through Revenue Quality, Deployment Depth and Cost Discipline

    Owners Can Improve Positioning Through Revenue Quality, Deployment Depth and Cost Discipline.

    Owners Can Improve Positioning Through Revenue Quality, Deployment Depth and Cost Discipline NeuraCap view · the questions this data puts on the table for the next twelve months Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. This page is Directional — NeuraCap views drawn from the analysis in this report. Observations, not recommendations. 15

    Everything on this page

    05 · STRATEGIC IMPLICATIONS Owners Can Improve Positioning Through Revenue Quality, Deployment Depth and Cost Discipline NeuraCap view · the questions this data puts on the table for the next twelve months Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. This page is Directional — NeuraCap views drawn from the analysis in this report. Observations, not recommendations. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 15 FOR OWNERS Choose the Model You Want the Market to Recognize Align product investment and capital allocation with a clear position across applications, infrastructure or managed technology services. Mixed models need disciplined pricing and cost ownership. FOR MANAGEMENT Turn Growth into Durable Production Revenue Focus execution on pilot-to-production conversion, retention, usage expansion and contracted demand. Growth carries more strategic weight when customers embed the product in critical workflows. FOR BOARDS Test Build-Versus-Buy Against Defensibility Evaluate whether proprietary data, distribution, workflow access or technical capability is faster to build or acquire. Keep integration demands and model dependencies central to the decision.

16. 16
    SECTION 06

    06

    06.

    Section 06 of 06 — The Full Universe, Methodology and Sources. Comparables detail behind every figure in the body, the valuation basis, and where each underlying disclosure lives. Use this divider to reset the room before the section's pages; the progress dots show where the argument stands. 16

    Everything on this page

    SECTION 06 06 APPENDIX The Full Universe, Methodology and Sources Comparables detail behind every figure in the body, the valuation basis, and where each underlying disclosure lives. 06 of 06 Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 16

17. 17
    06 · PUBLIC COMPARABLES (1 OF 2)

    Public Comparables on EV / Revenue (CY2027E), Grouped by Valuation Tier

    Public Comparables on EV / Revenue (CY2027E), Grouped by Valuation Tier.

    Public Comparables on EV / Revenue (CY2027E), Grouped by Valuation Tier Teal shading marks a EV/Revenue above the sector median (3.0x); amber marks below · 16 rated companies; 1 not rated (no eligible EV/Revenue) · tickers link to the underlying source · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. All 16 rated rows are in this appendix and in the companion workbook, which carries the complete field set. Names without an eligible multiple are listed in the companion workbook. 17

    Everything on this page

    06 · PUBLIC COMPARABLES (1 OF 2) Public Comparables on EV / Revenue (CY2027E), Grouped by Valuation Tier Teal shading marks a EV/Revenue above the sector median (3.0x); amber marks below · 16 rated companies; 1 not rated (no eligible EV/Revenue) · tickers link to the underlying source · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. All 16 rated rows are in this appendix and in the companion workbook, which carries the complete field set. Names without an eligible multiple are listed in the companion workbook. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 17 Company Ticker Segment EV EV/Revenue (CY2027E) Rev growth EBITDA margin Rule of 40 PREMIUM — ≥5.5x · median 8.2x · 4 companies Strategy Inc STRC AI-native enterprise applications $49.8B 97.7x 2% n/a n/a Samsara Inc. IOT AI infrastructure and model-enablement platforms $21.4B 8.9x 20% 24% 44 SoundHound AI, Inc. SOUN AI-native enterprise applications $2.3B 7.5x 30% n/a n/a BigBear.ai Holdings, Inc. BBAI AI implementation and managed technology services $944M 5.8x 12% n/a n/a CORE — 1.1x–5.5x · median 3.0x · 8 companies Palladyne AI Corp. PDYN AI infrastructure and model-enablement platforms $216M 5.4x 61% n/a n/a SentinelOne, Inc. S AI infrastructure and model-enablement platforms $7.1B 5.1x 18% 17% 35 C3.ai, Inc. AI AI-native enterprise applications $1.1B 4.6x 12% n/a n/a Innodata Inc. INOD AI implementation and managed technology services $2.1B 4.5x 28% 20% 47 Cerence Inc. CRNC AI-native enterprise applications $432M 1.5x 10% n/a n/a Rezolve AI PLC RZLV AI-native enterprise applications $759M 1.4x n/a 2% n/a Airship AI Holdings, Inc. AISP AI infrastructure and model-enablement platforms $50M 1.2x 48% 3% 51 PowerFleet, Inc. AIOT AI infrastructure and model-enablement platforms $624M 1.2x 11% 26% 37 DISCOUNT — <1.1x · median 0.7x · 4 companies Veritone, Inc. VERI AI-native enterprise applications $150M 0.9x 21% n/a n/a Gorilla Technology Group Inc. GRRR AI infrastructure and model-enablement platforms $568M 0.8x n/a 45% 273

18. 18
    06 · PUBLIC COMPARABLES (2 OF 2)

    Public Comparables on EV / Revenue (CY2027E), Grouped by Valuation Tier

    Public Comparables on EV / Revenue (CY2027E), Grouped by Valuation Tier.

    Public Comparables on EV / Revenue (CY2027E), Grouped by Valuation Tier Teal shading marks a EV/Revenue above the sector median (3.0x); amber marks below · 16 rated companies; 1 not rated (no eligible EV/Revenue) · tickers link to the underlying source · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. All 16 rated rows are in this appendix and in the companion workbook, which carries the complete field set. Names without an eligible multiple are listed in the companion workbook. 18

    Everything on this page

    06 · PUBLIC COMPARABLES (2 OF 2) Public Comparables on EV / Revenue (CY2027E), Grouped by Valuation Tier Teal shading marks a EV/Revenue above the sector median (3.0x); amber marks below · 16 rated companies; 1 not rated (no eligible EV/Revenue) · tickers link to the underlying source · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. All 16 rated rows are in this appendix and in the companion workbook, which carries the complete field set. Names without an eligible multiple are listed in the companion workbook. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 18 Company Ticker Segment EV EV/Revenue (CY2027E) Rev growth EBITDA margin Rule of 40 DISCOUNT — CONTINUED — <1.1x · median 0.7x · 4 companies SoundThinking, Inc. SSTI AI-native enterprise applications $62M 0.6x 12% 15% 27 Duos Technologies Group, Inc. DUOT AI-native enterprise applications $72M 0.5x n/a 8% 205

19. 19
    06 · PRECEDENT TRANSACTIONS (1 OF 2)

    All Precedent Transactions with Disclosed Terms, Newest First

    All Precedent Transactions with Disclosed Terms, Newest First.

    All Precedent Transactions with Disclosed Terms, Newest First 12 transactions with disclosed terms in this tier (70 recorded) · multiples on LTM financials at announcement where disclosed · deal values link to the underlying filing · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. 66 precedent record(s) carry data-quality flags (deal value unit unresolved; divestiture roles reassigned; duplicate precedent id); figures are shown as recorded in the filing. 58 recorded transactions with neither a disclosed value nor a multiple are omitted from the precedent list and the case studies (kept in the companion workbook). Deal multiples are LTM at announcement (Definitive, from filings); they are not directly comparable to the CY2027E public basis and no spread is claimed. 19

    Everything on this page

    06 · PRECEDENT TRANSACTIONS (1 OF 2) All Precedent Transactions with Disclosed Terms, Newest First 12 transactions with disclosed terms in this tier (70 recorded) · multiples on LTM financials at announcement where disclosed · deal values link to the underlying filing · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. 66 precedent record(s) carry data-quality flags (deal value unit unresolved; divestiture roles reassigned; duplicate precedent id); figures are shown as recorded in the filing. 58 recorded transactions with neither a disclosed value nor a multiple are omitted from the precedent list and the case studies (kept in the companion workbook). Deal multiples are LTM at announcement (Definitive, from filings); they are not directly comparable to the CY2027E public basis and no spread is claimed. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 19 Date Transaction EV EV/LTM Rev EV/LTM EBITDA Why it matters Nov-2025 Palladyne AI Corp. → GuideTech, LLC $6M n/a n/a Palladyne AI Corp. (PDYN) acquired GuideTech, LLC in a completed transaction recorded at $6M. The pairing provides a reference for smaller capability acquisitions. Sep-2025 Scilex Holding Company → Datavault AI Inc. $71M 15.9x n/a Scilex Holding Company completed its acquisition of Datavault AI Inc. The disclosed terms provide a high-end revenue reference within the transaction set. Jan-2025 n/a → Unusual Machines, Inc. n/a 2.3x 13.8x The completed transaction involving Unusual Machines, Inc. was recorded at 2.3x EV / Revenue and 13.8x EV / EBITDA. The two measures connect revenue scale with earnings delivery. Nov-2023 BigBear.ai Holdings, Inc. → Pangiam Intermediate Holdings, LLC $70M n/a n/a BigBear.ai Holdings, Inc. (BBAI) completed its acquisition of Pangiam Intermediate Holdings, LLC. The transaction adds a reference for strategic capability expansion. Feb-2022 Tantalus Systems Inc. → Congruititive n/a 2.0x 3.8x The announced transaction between Tantalus Systems Inc. and Congruititive was recorded at 2.0x EV / Revenue and 3.8x EV / EBITDA. The pricing offers a lower-range cross-check on both measures. Jun-2021 WiMi Hologram Cloud Inc. → VIYI Algorithm Inc. $400M n/a n/a WiMi Hologram Cloud Inc. announced a transaction involving VIYI Algorithm Inc. The disclosed value provides a larger reference within the set. Apr-2021 Microsoft Corporation → Nuance Communications, Inc. n/a 11.9x 59.7x Microsoft Corporation announced its acquisition of Nuance Communications, Inc. at 11.9x EV / Revenue and 59.7x EV / EBITDA. The pricing sits at the upper end of the disclosed transaction references. Feb-2021 AeroVironment → Intelligent Systems Group n/a 3.6x n/a AeroVironment announced its acquisition of Intelligent Systems Group at 3.6x EV / Revenue. The transaction offers a mid-range revenue reference. Feb-2020 Azenta Us Inc. → Ruro, Inc. n/a n/a 0.7x Azenta Us Inc. announced its acquisition of Ruro, Inc. with an EV / EBITDA reference. The transaction broadens the set beyond revenue-based pricing.

20. 20
    06 · PRECEDENT TRANSACTIONS (2 OF 2)

    All Precedent Transactions with Disclosed Terms, Newest First

    All Precedent Transactions with Disclosed Terms, Newest First.

    All Precedent Transactions with Disclosed Terms, Newest First 12 transactions with disclosed terms in this tier (70 recorded) · multiples on LTM financials at announcement where disclosed · deal values link to the underlying filing · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. 66 precedent record(s) carry data-quality flags (deal value unit unresolved; divestiture roles reassigned; duplicate precedent id); figures are shown as recorded in the filing. 58 recorded transactions with neither a disclosed value nor a multiple are omitted from the precedent list and the case studies (kept in the companion workbook). Deal multiples are LTM at announcement (Definitive, from filings); they are not directly comparable to the CY2027E public basis and no spread is claimed. 20

    Everything on this page

    06 · PRECEDENT TRANSACTIONS (2 OF 2) All Precedent Transactions with Disclosed Terms, Newest First 12 transactions with disclosed terms in this tier (70 recorded) · multiples on LTM financials at announcement where disclosed · deal values link to the underlying filing · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. 66 precedent record(s) carry data-quality flags (deal value unit unresolved; divestiture roles reassigned; duplicate precedent id); figures are shown as recorded in the filing. 58 recorded transactions with neither a disclosed value nor a multiple are omitted from the precedent list and the case studies (kept in the companion workbook). Deal multiples are LTM at announcement (Definitive, from filings); they are not directly comparable to the CY2027E public basis and no spread is claimed. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 20 Date Transaction EV EV/LTM Rev EV/LTM EBITDA Why it matters Feb-2019 Rekor Systems, Inc. → OpenALPR Technology, Inc. $5M n/a n/a Value shown as recorded in the filing; deal value unit unresolved. Dec-2018 Apis Ventures LLC → Veritone, Inc. $145M 7.4x n/a Value shown as recorded in the filing; deal value unit unresolved. Mar-2018 Motorola Solutions, Inc. → Avigilon Corporation n/a 16.5x 16.5x Value shown as recorded in the filing; deal value unit unresolved.

21. 21
    06 · METHODOLOGY

    Sources, Assumptions and Data Quality

    Sources, Assumptions and Data Quality.

    Sources, Assumptions and Data Quality How this report was built, what was excluded, and where every underlying disclosure lives · as of 2026-09-28 Every figure in this report links to the record it was taken from. Where a figure has no link, the appendix names its source and the basis on which it was read. 21

    Everything on this page

    06 · METHODOLOGY Sources, Assumptions and Data Quality How this report was built, what was excluded, and where every underlying disclosure lives · as of 2026-09-28 Every figure in this report links to the record it was taken from. Where a figure has no link, the appendix names its source and the basis on which it was read. Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice 21 VALUATION BASIS Primary valuation basis: EV / Revenue on CY2027E consensus (16 of 17 companies eligible; multiples failing the platform's plausibility gates are excluded, never plotted). Practitioners price this industry on EV / Revenue rather than EV / EBITDA; validated coverage supports the industry standard (16 of 17 companies), so this report follows it. EV / Revenue on CY2027E is the lead convention: it is the sector-appropriate prior for Artificial Intelligence Software and it clears the coverage gate with 16 of 17 companies (94%). P / E is carried as a cross-check. A revenue lens is used rather than a profit multiple because the set is not consistently profitable on a forward basis (9 of 17 names with a meaningful EBITDA). DATA QUALITY & EXCLUSIONS 64 records were excluded or quarantined by the platform's validation gates (unit errors, implausible ratios, ineligible securities and classification failures); the full ledger ships in the companion tables and never feeds a statistic in this report. DEFINITIVE VS DIRECTIONAL Definitive content is disclosed or consensus data passed through stated arithmetic. Directional content — tier boundaries, situation cuts, "why it matters" commentary, the agenda pages — is NeuraCap analyst judgment and is labeled as such where it appears. WHAT THIS REPORT IS NOT This report characterises a sector. It does not recommend buying, selling or holding any security, and no statement in it should be read as investment advice or a price target. WHERE THE DATA LIVES 735 registered source documents stand behind the figures in this report. Metric hyperlinks throughout the deck open the specific filing or data record; by publisher: sec.gov (734) · home.treasury.gov (1). The companion workbook beside this deck carries the complete source index.

22. 22

    The Premium Sits with Businesses That Pair Expansion with Credible Durability.

    The Premium Sits with Businesses That Pair Expansion with Credible Durability..

    Closing — The Premium Sits with Businesses That Pair Expansion with Credible Durability. The companion tables beside this deck carry the full universe, the exclusion ledger and the complete source index for any figure a client wants to trace. 22

    Everything on this page

    The Premium Sits with Businesses That Pair Expansion with Credible Durability. NeuraCap AI — Artificial Intelligence Software Coverage September 2026 · Prepared by NeuraCap AI · Confidential Artificial Intelligence Software Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 22

More coverage in Information Technology

• Sector Report · Sep 28, 2026Education Technology Software Sector Outlook — September 2026
• Sector Report · Sep 28, 2026E-Commerce Software and Platforms Sector Outlook — September 2026
• Sector Report · Sep 28, 2026Power Semiconductors and Power Electronics Sector Outlook — September 2026
• Sector Report · Sep 28, 2026Database and Data Infrastructure Sector Outlook — September 2026