AIM Research’s PeMa Quadrant report provides tech buyers with an analyst’s view of how the data science market is evolving and how vendors are adapting. The vendor profiles outline each firm’s core capabilities, and the PeMa Quadrant highlights their relative positions on market penetration and technology maturity. While the report and the Quadrant acts as a comprehensive guide for tech buyers, it should be treated as one leg of the stool among other RFP criteria (your buying process), since each tech buyer’s maturity varies and they will require a partner suited to their specific capability gaps.
The large-scale data science services market is undergoing structural repositioning as the traditional basis of competition, delivery scale, cost arbitrage, technology breadth, erodes against enterprises that now carry significant internal AI capability, operate distributed GCCs, and hold direct hyperscaler relationships. The value question has shifted from doing data science on behalf of clients to industrializing, governing, and continuously operating analytics at a scale clients cannot replicate internally; engagements won previously on execution now require demonstrated analytical judgment, proprietary accelerator depth, and production-scale outcome evidence, with engagement models diversifying accordingly toward outcome-linked and agentic-component-based constructs. Traditional ML retains centrality in the highest-value use cases including forecasting, pricing, risk, and fraud, while GenAI expands scope by accelerating discovery and incorporating unstructured data into decision systems, operating alongside rather than replacing it. The most consequential shift is the convergence of GenAI, Agentic AI, and ML into integrated delivery systems: agentic architectures are changing how data science is designed and operated, with use-case discovery, governance monitoring, and analytical narrative generation increasingly automated within pipelines, and multi-agent systems beginning to replace single-model deployments in complex workflows. Reinvention is visible across provider landscape: advisory-led firms repositioning around AI risk and responsible AI operationalization, IT services-led firms industrializing through platform investment and hyperscaler co-builds, and domain-embedded firms moving toward outcome-linked structures where analytics directly governs measurable process metrics.
Process reimagination is emerging as the most ambitious work pattern. A subset of providers is moving beyond automating existing processes toward redesigning them with AI built in from the ground up, most visibly in insurance claims, financial operations, field service management, and clinical workflow support.
Work is predominantly embedded, not standalone. Across all fifteen providers, data science is delivered as a component of larger transformation programs (cloud modernization, managed services, or operational process redesign) rather than as independently scoped analytics engagements. Legacy warehouse migration to cloud-native lakehouses across hyperscaler and specialist data platforms is the single most commercially active workstream, serving as the upstream foundation before modeling work begins.
Regulated industry programs dominate the highest-value work. Financial crime detection, credit and risk modeling, clinical analytics, actuarial modeling, regulatory compliance automation, and underwriting intelligence account for the largest share of evidenced, production-scale data science delivery. These use cases require model governance, auditability, and domain expertise that create durable barriers to entry.
IT-OT convergence is a distinct and growing work category. Providers with engineering heritage are integrating sensor, edge, and operational technology data with IT analytics platforms, creating a category of industrial data science covering predictive maintenance, digital twin, quality inspection, and grid analytics that general-purpose IT services providers cannot easily replicate.
GenAI is applied to unstructured data problems, not replacing ML. The clearest production GenAI pattern is document and knowledge intelligence using RAG architectures across regulatory documents, engineering knowledge bases, clinical records, and financial filings. Traditional ML continues to govern forecasting, risk, supply chain, and fraud use cases.
Agentic AI is entering operational workflows. Multi-agent systems are appearing in financial operations, supply chain coordination, industrial maintenance, and field service management, moving from pilot to production in a subset of providers.
A total of 15 vendors are featured in the PeMa Quadrant study Vendors are evaluated on delivery scale and financial health, growth, customer confidence, and company outreach, which reflect market penetration (Pe), and on work delivery, tech advancement, employee maturity, and support infrastructure, which reflect technology maturity (Ma).
Featured Vendors (in alphabetical order):
Accenture, Capgemini, CGI, DXC, EY, Fujitsu, Genpact, HCLTech, Hitachi Digital Services, Infosys, Kyndryl, NTT DATA, TCS, Tech Mahindra, and Wipro
Table of Contents:
Market Outlook:
The Data Science Services PeMa Quadrant:
Vendor Profiles – Capabilities and Differentiators
How to access the report?
To access the full report, you may purchase it for USD 10,000 for internal use. A separate reprint license is required for any external or marketing use. Please reach out to info@aimresearch.co for further details on the commercials.
$10,000.00
