Competitive Intelligence · October 2026

Five patterns shaping the
data → AI → CE transition

How 15 IT services firms are repositioning from data analytics companies to AI companies — and where context engineering fits in each play.

15 firms profiled across 3 research tracks
$3.3B Capgemini paid for WNS (2.6× revenue) — CE-adjacent IP premium
12–24 months before Tredence or Tiger claims CE analyst recognition
0 firms have named a CE practice on Microsoft Fabric
Five winning patterns
01
Platform IP Build
Own the product layer — CE as the product's semantic foundation
The most capital-intensive path: build a named, branded AI platform that the firm can sell and license independently of headcount. Every successful platform in this set shares one trait — a CE-equivalent layer underneath, whether called a semantic layer, knowledge graph, or context architecture. The platforms that go to market without one (e.g., rebranded dashboards) fail to retain enterprise clients past the pilot.
Fractal · Cogentiq + Fathom-R1 IPO 2.66× Persistent · iAURA 2.0 + Brickbuilder BFSI +16% YoY Quantiphi · baioniq VPC-native Everest Star Performer EPAM · DIAL open-source $5.3B rev, +14%
CE role → product's semantic architecture; agents ground via the platform's context layer
02
Ecosystem Anchor
Become the hyperscaler's CE partner of record — co-sell drives pipeline
Pick one hyperscaler's CE stack and become the #1 partner in it. The Databricks Elite tier + annual Partner of the Year award is the most visible signal: it puts the SI in the vendor's co-sell motion and earns it shelf space in the Genie/Unity Catalog ecosystem. The firms that win multiple consecutive years (Quantiphi/Google, Tiger/Databricks, NTT Data/Snowflake 5× APJ) compoundly improve their deal conversion because the hyperscaler's own sales team starts bringing them in. CE certification is the credential that unlocks the top tier.
Tiger Analytics · Databricks Elite + 2025 EAPY $350M → $1B target Quantiphi · Google Cloud 4× Partner of Year Gartner Emerging Leader NTT Data · Microsoft GSGC + Snowflake Elite + Databricks $200M $30B rev ZS Associates · Premier Databricks + Premier Snowflake Everest Leader Life Sciences 3×
CE role → Databricks CE cert is the access credential; co-sell requires it
03
Semantic First
Claim CE vocabulary before the market solidifies — first mover on the label
A small number of firms have begun claiming the CE vocabulary explicitly — in analyst publications, product names, or regulatory filings — while larger peers still use proprietary language. ThoughtWorks coined data mesh in 2019 and built a decade of thought leadership from it; the same window is open for CE in 2026. The firms making this move are doing so through publications and frameworks, not product launches — which means the moat is still buildable without a $50M R&D budget.
Tredence · "decision catalog" + Atlan joint CE demo Forrester Wave #1 Leader ThoughtWorks · Tech Radar "semantic layer as Trial" + CE article Forrester AI Services Leader ZS Associates · AIESC product (4-layer semantic model on AWS) AIESC named on AWS Marketplace Coforge · "Intent Engineering" BSE exchange filing Oct 2026 First BSE-filed CE framework
CE role → CE is the category name, not just a technique; own the vocabulary first
04
Vertical + Regulatory Moat
Own one industry's CE layer — compliance framing locks in the budget
The fastest-growing firms in this set are not the most horizontally ambitious — they're the ones that went deep into one vertical and connected their AI/CE layer directly to a regulatory mandate. EXL's insurance/F&A CE work is framed as operational resilience. ZS's AIESC is built on pharma commercial data standards. Persistent's iAURA BFSI frames CE as the mechanism for BCBS 239 lineage. Compliance doesn't just accelerate the deal — it survives budget cuts and makes the SI indispensable rather than interchangeable.
EXL Service · Insurance/F&A CE, 60% Data-AI rev +28–31% YoY Data-AI segment ZS Associates · AIESC pharma semantic stack, AWS Marketplace Everest Life Sciences Leader 3× Persistent · iAURA BFSI + BCBS 239 alignment BFSI revenue +16.2% YoY Genpact · F&A/supply chain "industry-specific semantic layer" HFS Market Leader Insurance
CE role → CE delivers the compliance evidence (BCBS 239 lineage, FHIR, FIBO); regulator is the forcing function
05
Acquisition
Buy CE capability rather than build it — compress the timeline to 12–24 months
Building a CE practice organically takes 3–5 years to reach analyst recognition. Acquisition compresses this to 12–24 months — and the exit multiples confirm the market is pricing CE-adjacent IP at a premium. The Capgemini/WNS deal is the clearest evidence: WNS had flat revenue growth yet commanded 2.6× revenue because of domain AI in F&A and insurance. LatentView's Beagle GPT acquisition directly drove +23% revenue growth in the following quarter. Tredence's KMK acquisition added pharma/life sciences coverage in a single step.
WNS → Capgemini $3.3B 2.6× revenue multiple LatentView → Decision Point/Beagle GPT $39.1M +23% revenue next quarter Tredence → KMK Consulting (pharma) Pharma → 25% of rev by 2028 NTT Data → WinWire (Fabric/Azure AI Foundry) May 2026
CE role → acquire a firm with domain ontology + CE delivery track record; IP survives the integration
The failure pattern
✕
The Label Swap
Rename the product. Don't rebuild the company.
Mu Sigma is the clearest case study: "decision sciences" became "AI-enabled decision intelligence" with new product names (muAoPS, Akashic Architecture, muTalos). The vocabulary changed. The platform didn't. No hyperscaler anchor was established. No named AI product gained analyst coverage. No CE or semantic layer work appeared in any public material. The result: India entity revenue fell 16.1% in FY24 (₹857.7Cr → ₹719.5Cr), headcount declined ~6%, and the firm dropped out of every major analyst recognition list in 2024–2026.

The pattern recurs in smaller ways across firms that rebrand data services as "AI-powered" without building the underlying context layer — initial client interest followed by high churn when pilots don't scale.
Revenue −16.1% FY24 Zero Gartner / ISG / Everest recognition 2024–2026 No hyperscaler partnership tier No named CE/AI platform Headcount shrinking −6% YoY
Company landscape at a glance

Sorted by estimated revenue (ascending). Growth figures are approximate; private company data is estimated from secondary sources.

Firm Revenue (est.) 3–5yr Growth Trend Primary Domain Cloud / Hyperscaler Focus CE Pattern Analyst Recognition
Sigmoid ~$75M est. ~15–20% est. CAGR (limited public data) Retail / CPG · BFSI Databricks · Snowflake Semantic First Not yet covered
Mu Sigma ~$86M FY24 −16.1% FY24 (declining) Retail · BFSI · Insurance No prominent hyperscaler partnership Label Swap ✕ Zero 2024–2026 coverage
LatentView ~$125M FY26 +23% after Decision Point acquisition BFSI · Technology · Analytics Databricks · Snowflake Acquisition NSE listed
the SI ~$180M est. ~15–20% est. CAGR; CE practice forming Data Engineering · Analytics · BFSI · Healthcare Databricks · Microsoft Azure · AWS Semantic First Not yet covered (opportunity)
Quantiphi ~$200M+ est. ~35–40%+ CAGR (high-growth phase) Healthcare · Retail · Media / Entertainment Google Cloud 4× Partner of Year · Databricks Ecosystem Anchor Gartner Emerging Leader, Everest SP
Fractal Analytics ~$290M FY25 ~20–25% CAGR; IPO pipeline active BFSI · CPG · Insurance Azure · Google Cloud · AWS Platform Build Forrester Leader, Everest Leader
Tiger Analytics ~$350M 2024 ~30%+ CAGR Retail / CPG · BFSI · Healthcare Databricks Elite (EAPY 2025) · OpenAI Select Ecosystem Anchor Everest Leader, ISG Leader
Tredence ~$350M 2025 ~40%+ CAGR (VC-backed growth phase) Retail / CPG · BFSI Databricks Elite · Snowflake · Atlan Semantic First Forrester Wave #1 Leader
ThoughtWorks ~$1.0B TTM 2024 −11% FY24; restructuring underway Tech Consulting · Retail · BFSI Multi-cloud (AWS · GCP · Azure) Semantic First Forrester Leader AI Services
Persistent Systems ~$1.1B FY25 +30%+ YoY (recent); BFSI vertical +16.2% BFSI · Healthcare · Technology Databricks Brickbuilder BFSI · AWS · Azure Vertical + Regulatory Databricks Brickbuilder BFSI
ZS Associates ~$1.5B est. ~15–20% est. CAGR (private) Pharma / Life Sciences · Healthcare Databricks Premier · Snowflake Premier · AWS Marketplace (AIESC) Ecosystem Anchor Semantic First Everest Leader Life Sciences 3×
EXL Service $2.09B FY25 +15% overall CAGR; Data-AI segment +28–31% YoY Insurance · BFSI · Healthcare Azure · Databricks · Snowflake Vertical + Regulatory Everest Leader, HFS Leader
Genpact ~$5.0B 2024 +8–10% overall; Data-Tech-AI segment +20%+ F&A · Insurance · Supply Chain · Healthcare Azure (Cora AI platform) · Google · Salesforce Vertical + Regulatory HFS Leader, Everest Leader
EPAM Systems ~$5.3B LTM 2025 +9% recovering (Ukraine revenue impact 2022–23) Tech / Software Eng. · BFSI · Healthcare Multi-cloud (AWS · Azure · GCP); DIAL open-source Platform Build Forrester Leader App Dev
NTT Data ~$30B ~8–10% org. CAGR; inorganic via WinWire (2026) BFSI · Telco · Manufacturing · Public Sector Microsoft GSGC 2025 · Snowflake Elite · Databricks $200M partnership Ecosystem Anchor Microsoft GSGC 2025
CE Pattern legend
Platform Build Built a named AI product sold independently of delivery headcount. CE sits as the semantic foundation underneath it.
Ecosystem Anchor Became the preferred CE/AI partner within a hyperscaler's ecosystem (Databricks Elite, Google Cloud PoY, Microsoft GSGC). Co-sell motion drives pipeline.
Semantic First Claimed CE vocabulary explicitly — in analyst publications, product names, tech radar entries, or filings — before the category has a dominant owner.
Vertical + Regulatory Owns one industry's CE layer and ties it to a compliance mandate (BCBS 239, FHIR, EU AI Act). Regulatory framing makes the engagement indispensable rather than interchangeable.
Acquisition Acquired CE-adjacent IP or domain capability instead of building it — compressing the practice timeline from 3–5 years to 12–24 months.
Label Swap ✕ Rebranded existing services as AI without platform investment, hyperscaler anchor, or CE depth. Revenue declined as clients recognised the gap between claim and delivery.
What this means for the SI
The category has no owner. The window is 12–24 months.
Across 15 firms and three research tracks, few have formally claimed the CE category on Microsoft Fabric or Azure AI Foundry. Tredence is closest (Semantic First pattern, Atlan partnership), but their positioning is still methodology-level. A firm that combines ecosystem anchor + semantic vocabulary claim + vertical regulatory framing in a coherent package would have a meaningful early-mover position.
Patterns to adopt — and which to skip
✓ Adopt these
Ecosystem Anchor — Databricks CE cert cohort is the fastest path to Elite co-sell. Microsoft co-sell follows the same logic on Fabric.

Semantic First — Claim CE vocabulary explicitly on Fabric and in BFSI/healthcare compliance framing. ThoughtWorks coined data mesh in 2019 and spent a decade on it. This window is shorter.

Vertical + Regulatory Moat — Pick one industry (BFSI or healthcare), map CE deliverables to its specific regulation, and make that the packaged accelerator. Compliance framing locks in the budget and makes the engagement indispensable.
✗ Skip or defer
Platform IP Build — Requires $20–50M R&D and 3–5 years. Fractal and Persistent spent this on iAURA and Cogentiq. the SI can partner for platform instead of building.

Acquisition — Capital-intensive and execution-heavy. The right play after CE practice is established and there's a specific IP gap to fill (e.g., a BFSI ontology vendor or a FHIR specialist).

Label Swap — The Mu Sigma failure pattern. Renaming "data analytics" as "AI-powered CE" without building the underlying platform or earning analyst recognition generates zero durable advantage.
Which industries are spending most in AI — and why CE wins the budget
Global enterprise AI services spend 2026 (est.). CE accelerators capture the governed-AI share of this — the deployment budget, not the model budget.
Financial Services
~$45B
BCBS 239 enforcement, EU AI Act Annex III mandatory, model risk governance (SR 11-7). 29 G-SIBs spending; CDO + CRO are economic buyers.
Healthcare & Pharma
~$35B
CMS FHIR prior auth Jan 2027 deadline, ONC HTI-1 AI risk disclosure rule. UnitedHealth alone: $3B AI budget 2026–2027. NHS: £900M. Healthcare KG market 16% CAGR.
Retail & CPG
~$25B
Commercial urgency (no regulation), but "missing semantic layer" is the buyer's own vocabulary for the CE problem. Carrefour: €100M/yr target. Market crowded — differentiate on GS1 + Genie Space.
Manufacturing
~$20B
Siemens already has a commercial KG product with pre-sales hiring. Predictive maintenance, supply chain ontology, digital twin context layers. Under-served by Indian SIs today.
Energy & Utilities
~$12B
CSRD reporting (EU), carbon accounting ontologies, grid optimization context layers. Early-stage CE buying — first-mover advantage available. EU Taxonomy regulation creates compliance framing.
Industry use cases to build — by pattern × vertical
Vertical Pattern Use case Platform Deal size
BFSI Vertical + Regulatory BCBS 239 CE compliance accelerator — maps Unity Catalog lineage to all 14 principles with FIBO alignment. CDO gets a regulator-ready evidence pack. Databricks + Fabric $300K–1.2M
BFSI Semantic First EU AI Act Annex III explainability layer — structured context delivery for credit scoring and fraud models, producing audit evidence by design. Fabric + Azure AI Foundry $150K–600K
Healthcare Vertical + Regulatory FHIR R4 → OMOP CDM context accelerator — payer prior auth API compliance by Jan 2027; SNOMED/LOINC normalization + MCP server for clinical AI agents. Azure Health Data Services + Databricks $400K–1.5M
Retail / CPG Ecosystem Anchor GS1-native demand signal context layer — product ontology + RGM semantic model delivered as a Databricks Genie Space; co-sell through Databricks Elite channel. Databricks Genie $200K–800K
Manufacturing White Space Digital twin context layer — supply chain entity graph + predictive maintenance ontology. Snowflake Cortex Analyst or Fabric Copilot as the query surface. No Indian SI owns this. Snowflake + Fabric $250K–900K
Tech spaces to invest in — ranked by the SI' leverage
①
Databricks CE certification + Unity Catalog governance
Cost: $200/person, 3 months part-time study. Return: Elite tier co-sell, $5–10M pipeline per certified cohort. No competing Indian SI has announced a cohort. The cert covers Unity Catalog, AI Search, MCP integrations, multi-agent context, and context window management — the complete CE delivery stack.
②
Microsoft Fabric IQ + Semantic Models + Azure AI Foundry
An underused CE substrate for Microsoft-anchor enterprises. Every Power BI customer already has a semantic model that can be converted into a CE asset in a 6-week engagement. Fabric IQ + MCP + Azure Health Data Services forms a coherent FHIR compliance stack. A published reference architecture naming this a CE platform would give the SI a concrete Fabric CE story in the market.
③
Neo4j + GraphRAG for domain knowledge graphs
Neo4j GraphAcademy has free certification paths. GraphRAG (retrieval over a knowledge graph instead of flat vector search) delivers 3× accuracy improvements in production (per Cognizant and Microsoft Research data). For BFSI (FIBO), healthcare (FHIR ontology), and manufacturing (asset hierarchy), a Neo4j knowledge graph is the semantic substrate every other accelerator runs on.
④
MCP (Model Context Protocol) connector engineering
MCP is now table-stakes for every enterprise AI agent framework. The scarcest skill is building pre-certified MCP context connectors for vertical domain data — FHIR clinical data, FIBO financial entities, SAP ERP entities. Corvera built MCP-native CPG connectors and signed 12 brands in 5 weeks. An SI that ships 5 pre-built domain MCP servers owns the context delivery layer for its verticals.
⑤
Atlan CE control plane as platform-agnostic differentiator
Atlan sits above Databricks, Snowflake, and Fabric simultaneously. A the SI CE practice built around Atlan as the metadata and governance layer is platform-agnostic — which matters for the 40% of enterprises running multi-cloud. Tredence already has an Atlan joint demo. The move is to go deeper: build a named BFSI or healthcare CE accelerator on Atlan rather than just referencing it.
The mental model shift — from delivery shop to CE product company
Current mental model
"We deliver data engineering projects on Databricks/Snowflake/Fabric."
"Our value is implementation capacity."
"We're measured by hours delivered and certifications held."
"We position against other Indian SIs on price and team size."
"AI is the client's strategy; we execute their roadmap."
CE product company mental model
"We own the CE layer that makes enterprise AI work on any cloud."
"Our value is the IP we bring, not the hours we sell."
"We're measured by AI accuracy improvement in client production."
"We position against compliance risk and AI failure — not other SIs."
"We bring the AI strategy through the CE accelerator we pre-built."
1
Build a certified cohort — Databricks CE Associate is GA (July 2026, $200/exam). A named certified cohort constitutes a practice. CE cert cohorts are still rare among mid-size SIs, and certification evidence is increasingly relevant to Databricks Elite co-sell conversations.
2
Publish a Fabric CE reference architecture — a document naming Fabric IQ + Semantic Models + Azure AI Foundry + MCP as an enterprise CE stack. This gives the SI a concrete technical story for Fabric-anchor accounts and a basis for Microsoft co-sell alignment. Cost: 2 weeks of architect time.
3
Launch the BCBS 239 CE Compliance Accelerator — a packaged offering that maps CE deliverables to the 14 BCBS 239 principles with FIBO alignment. The CDO gets a regulator-ready evidence pack. Only 2 of 31 G-SIBs pass all 14 principles. ECB has named it the #2 supervisory priority through 2027. The first SI to package this wins the BFSI CE lane.
4
Ship a named domain MCP context server — one pre-built MCP server for FHIR clinical data or FIBO financial entities, published open-source. The distribution is every LLM framework integrating MCP. Corvera got 12 CPG brand clients in 5 weeks with an MCP-native product. the SI can replicate this in BFSI or healthcare with domain depth no CPG startup has.