Prasenjeet Jena
Product Manager @NielsenIQ
Signup · Get unlimited contacts
WORK HISTORY
Product Manager @NielsenIQ
Chennai, IN
Strategy & Roadmap: Defined roadmap for enterprise AI capabilities spanning GenAI services, evaluation frameworks, and scalable platform architecture aligned with business and client priorities.GenAI Platform Leadership: Led 0→1 development of enterprise GenAI services including “Arthur,” a scalable AI assistant adopted across global to enhance analytics workflows and increase platform engagement.Auto-Report Configuration: Developed AI-assisted data selection and querying capabilities to accelerate report configuration by 70% and enable business users to self-serve insights.AI Evaluation & Trust: Built AI evaluation frameworks incorporating hallucination detection, scoring mechanisms, and guardrails to improve reliability and production readiness of LLM-powered solutions with accuracy of 85% and precision of 82%.RAG Infrastructure: Led Retrieval-Augmented Generation (RAG) pipelines to ground AI responses on proprietary documents, improving response accuracy while optimizing latency and cost efficiency.Translation Automation: Launched AI-driven translation services leveraging RAG to streamline multilingual content generation while reducing external dependency and cost.Workflow Automation: Implemented AI-powered ticket deflection and resolution workflows, reducing manual support effort and improving self-service efficiency across internal and client-facing operations.Multi-agent framework & Governance: Working on multi-agent governance principles including design, deployment, entitlement controls, audit traceability, evaluation workflows, and usage guardrails to support enterprise use-cases.
EDUCATION
Siksha 'O' Anusandhan University
Bachelor of Technology - BTech, Information Technology
Golden Gate University
Master of Business Administration - MBA, Business Analytics
ABOUT PRASENJEET JENA
Most AI features get built. Few get trusted.That\'s the gap I have spent the last 3 years living in. But getting here took 7.5+ years of building products across very different problems.I started as a developer intern, building chatbot components in Python. That hands-on technical grounding pushed me into product — first at Racetrack.ai, where I shipped a self-serve CMS that cut client onboarding from 2-3 weeks to 3-5 days.At Muvi.com, the challenge shifted to scale. I ran a full build-vs-buy evaluation for AI recommendations — ruled out Amazon Personalize and Recombee as cost-prohibitive for our SMB client base — and shipped an in-house recommendation engine deployed across 20+ OTT platforms that increased watch time by 35%.Nokia was a different kind of test — coordinating a zero-downtime global rollout to employees across 20+ offices for desk booking. Less product invention, more execution discipline.NielsenIQ is where all of it came together and where I grew the most.I joined leading a single product team on the core platform: launched report scheduling, drove a full UI redesign, introduced advanced filtering, and improved asset sharing rules across the Discover platform. Foundational work, but it taught me how enterprise users actually behave when a product is their daily tool.From there I was pulled into the early GenAI team - starting with RAG implementation, figuring out how to make retrieval actually work for NielsenIQ\'s domain-specific data. That work grew in scope until I was leading the full GenAI ART, coordinating ~35 AI engineers across three scrum teams, owning cross-team dependencies and roadmap sequencing for the entire AI platform. The work isn\'t glamorous in the way AI gets written about. It\'s decisions like: do we build our own translation pipeline or pay $80K a year to vendor and accept brand compliance failures? How do we know when the model is confident enough to trust? How do we give 100k users an AI that speeds them up without becoming a liability when it\'s wrong?Those questions and not the just model benchmarks is what enterprise AI actually runs on. Outside of my core work, I am building AI agent projects end-to-end using Agentic IDEs, LangGraph, Azure AI Foundry, MCP, and Google ADK. Not to become an engineer. To stay honest about what I am asking engineering teams to build, and to keep learning faster as the field moves. Always interested in conversations about enterprise AI, multi-agent systems, and what it actually takes to operationalise LLMs at scale. https://github.com/prasenjeet-jena
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.