Harshvardhan Sharma

About Harshvardhan Sharma

I architect systems that bridge the gap between static models and autonomous enterprise agents.With over 13 years of experience across the AI/ML lifecycle, my focus has shifted from core predictive modeling to the orchestration of complex, stateful Generative AI ecosystems. I specialize in the \"frontier\" problems of modern AI: managing long-term context, implementing persistent memory layers, and designing multi-agent workflows that can plan and execute across fragmented enterprise data silos.At Tata Consultancy Services, I lead the design of intelligent architectures that transform how global enterprises interact with their data. Whether it is building Natural Language Interfaces for specialized scientific research or engineering self-healing support ecosystems within ServiceNow, my approach is rooted in creating scalable, production-ready solutions that move beyond the \"PoC\" phase.I believe the next era of enterprise value won\'t come from larger models, but from more sophisticated orchestration. I spend my time solving for state management, episodic memory, and the seamless integration of GenAI, NLP, and Core ML to build a unified \"corporate brain.\"Core Areas of Focus:• Agentic Orchestration: Designing collaborative multi-agent systems for complex, multi-step reasoning.• Context & Memory Engineering: Building persistent state layers that allow AI to maintain personalization and relevance across long-term interactions.• Scalable AI Solutions: Engineering end-to-end pipelines that integrate Generative AI with traditional ML and NLP to solve high-stakes business challenges.• Unified Data Intelligence: Creating architectures that treat disparate platforms (M365, ServiceNow, SQL) as a single, searchable pool of knowledge.

Experience

1. Solution Architect -data Science

   Tata Consultancy Services

   Jan 2023 — Present · IN

   Intelligent Support Ecosystems: Architected a GenAI-driven orchestration layer for ServiceNow to automate complex ticket triaging and resolution workflows.Natural Language Interfaces (NLI): Designed and implemented a conversational interface for Laboratory Information Systems, enabling researchers to query complex scientific databases using natural language.Agentic Test Automation: Engineered an LLM-based pipeline that autonomously transforms unstructured user stories into structured functional test suites, improving requirement traceability.Enterprise RAG Strategy: Developed scalable retrieval-augmented generation frameworks that unify cross-platform data for seamless organizational knowledge access.

Education

• Rajasthan Technical University, Kota

  Bachelor of Technology - BTech, Computer Science