The model is a commodity
Same model, radically different outcomes. What separates them is system design, data quality, retrieval architecture and the shape of the context you hand it — not the frontier release on the label.
About
Abhishek Prasar
Enterprise AI Transformation Leader, WNS Global Services (now Capgemini) · Gurugram, India
I have spent fifteen years building AI inside enterprises — first the models, then the platforms around them, and now the organisations that have to live with both. The journey has tracked the field itself: from classical machine learning on enterprise data, to natural language and people analytics at scale, to generative AI, and now agentic systems in production.
The tools changed beyond recognition over those years. The problem underneath never moved. Whether the system was a classical model a decade ago or a multi-agent one today, the questions that decided everything were the same: whose data is this, who is allowed to see the answer, how do we know when it is wrong, and will anyone actually use it. The model at the centre was rarely the hard part. Everything around it is where the work lives.
Today my work sits with the commercial function — designing the architecture, the governance, and the adoption that turn AI from a promising demo into something an enterprise runs on. I have built AI capabilities from the ground up and delivered platforms used across global organisations, and the lesson has been consistent at every scale: capability is not the constraint. Trust, data boundaries, and adoption are.
That is the half of the job I write about here — the unglamorous half that no benchmark measures and every production system depends on. If the model is the easy part, this is the rest of it.
How I think
Three things I keep coming back to
Not a philosophy. Just the patterns that have held up across enough organisations, and enough generations of the technology, to trust.
Governance from day one
Data boundaries, DLP, PII handling and responsible-AI review are cheap to design in and expensive to retrofit. In regulated environments they are also the difference between a pilot and something that ships.
Adoption is the product
A platform nobody opens has delivered nothing. Most resistance is rational — people burned by earlier tools, missing access, misaligned incentives — and each of those needs a different intervention.
Background
Where I have worked
The short version. The longer one is on LinkedIn.
- Sept 2025 — present
Enterprise AI & Sales Transformation Leader
WNS Global Services (now Capgemini) · Gurugram
Enabling the commercial organisation with AI — owning the architecture, governance and adoption for in-house, enterprise-governed AI tools built to sit safely inside the business. Focused on lifting sales efficiency and cutting the manual effort sellers spend away from selling, with adoption treated as a first-class outcome rather than an afterthought. Support the sales function across every vertical, turning agentic systems from promising pilots into capability the commercial teams rely on day to day.
- Nov 2019 — Sept 2025
AI/ML Leader — GenAI & Enterprise Intelligence
LTIMindtree · Bengaluru
Founded and led the AI/ML practice — building it from the ground up into the delivery engine for enterprise AI across the organisation, and becoming the face of the AI team to the CIO office. Stood up the first team to bring generative AI into production, ahead of the market and ahead of the mandate, and set the direction that the wider organisation later followed. Delivered across the enterprise’s core functions — a talent and HR transformation that put AI in the hands of over a hundred thousand employees, and sales enablement that gave commercial teams AI-drafted proposals and faster pursuit. Championed enterprise Copilot adoption into the core infrastructure, and collaborated with technology partners such as Snowflake and Microsoft to shape early product direction. Throughout, replaced black-box vendor tools with explainable, in-house platforms the business could own and govern — work that earned two CIO awards for its impact. The pattern across six years: see the shift early, build it in-house, and put it in production while others were still evaluating.
- Mar 2019 — Oct 2019
Data Science & Analytics Lead
Dell Technologies · Bengaluru
Led analytics initiatives across the supply chain function, turning operational data into decisions leadership could act on. Built a multilingual product-feedback pipeline that translated customer feedback from languages across global markets into English, then applied NLP to route each issue into the right category at around 90% accuracy — layering sentiment analysis on top to separate signal from noise. Delivered the results as executive Tableau reporting that gave leadership a live view of product issues as they emerged. Worked alongside global partners to shape the analytics strategy behind it — moving the function from scattered reporting toward a single, trusted decision surface.
- Feb 2018 — Feb 2019
Lead Data Scientist
Tata Consultancy Services · Bengaluru
Led a data science team supporting top-tier global clients across telecom and banking, building and maintaining the credit-risk models their lending decisions depended on. Applied NLP for text classification and entity extraction on enterprise document estates, and brought structure to unstructured client data at scale — sharpening both the accuracy and the speed of the insight that reached the business.
- Apr 2015 — Jan 2018
Machine Learning Engineer
Ericsson · Gurugram
Built machine learning models for one of telecom’s most expensive problems — subscriber churn — predicting which high-value customers were likely to leave in time for the business to act, and helping protect the revenue base that retention teams fight to keep. Also delivered models for network performance analytics, along with the production Python pipelines behind them, serving operator clients across the USA and South Africa.
- 2011 — 2015
Data Scientist / ML Engineer
Saburi TLC · Axon Network Solutions · ASMA Financial Solutions
Predictive analytics across telecom, networking and financial services — where the foundations were laid.
Credentials
Education and the stack I build on
Tools change every eighteen months. These are the ones currently in hand.
- MBA Indian Institute of Management (IIM) Indore
- B.Tech, Computer Science
How I work
Strategy & Ownership
Governance & Risk
Agentic & GenAI Architecture
Machine Learning
Platforms
Working on the same problems?
I am glad to hear from AI and engineering leaders, founders, and anyone who disagrees with something I have written here.