Hi, I'm Hari Jaiswal
AI Transformation Lead | Multi-Agent Systems | Enterprise AI Strategy
I help organizations transform through AI -- from strategy and maturity assessment to building multi-agent systems and scaling enterprise-wide adoption.
About Me
Hello! I'm an AI Transformation Leader with 20+ years in enterprise IT.
My career has spanned the full arc of enterprise technology -- from hands-on development at Microsoft, to leading Agile transformations at Deutsche Bank, to now driving AI adoption at scale for UHG Optum through TCS.
Currently, I lead AI transformation initiatives at TCS for UHG Optum, where I design multi-agent systems, conduct AI maturity assessments, and build training programs that help enterprise teams move from AI curiosity to AI competency.
What sets me apart is the combination of deep technical skills in AI/ML (M.Tech from BITS Pilani, hands-on research in RL and multi-agent systems) with two decades of experience navigating enterprise complexity -- stakeholder management, change management, and large-scale program delivery.
I've worked across Healthcare (UHG/Optum), Banking & Finance (Deutsche Bank, Russell Investments, Vanguard), and Hi-Tech (Microsoft) -- bringing cross-domain perspective to every AI initiative I lead.
Career Journey
20+ years of building, leading, and transforming
AI Transformation Lead
TCS / UHG Optum
Sep 2025 – Present
Leading enterprise AI transformation -- designing multi-agent systems, conducting AI maturity assessments, building AI training programs, and driving organization-wide AI adoption strategy.
AI and Data Engineering Lead
TCS / UHG Optum
Nov 2023 – Aug 2025
Led AI/ML engineering initiatives including RAG-based chatbots, MLOps pipelines, and data engineering solutions for healthcare analytics. Pursued M.Tech in AI & ML from BITS Pilani alongside.
Enterprise Transformation Coach
TCS / Deutsche Bank
Apr 2019 – Oct 2023
Coached 20+ teams through Agile transformation at Deutsche Bank. Designed and delivered transformation frameworks, established communities of practice, and drove continuous improvement culture.
Scrum Master / Agile Project Manager
TCS / Russell Investments, Vanguard
Mar 2011 – Mar 2019
Managed Agile delivery across financial services -- leading Scrum teams, facilitating SAFe ceremonies, and ensuring on-time delivery for investment management platforms.
Development Lead
TCS / Microsoft
Oct 2006 – Feb 2011
Led development teams building enterprise solutions on the Microsoft technology stack. Hands-on coding, architecture decisions, and team mentorship in a fast-paced product environment.
Prior Roles
Nihilent Technologies, AESSeal India, Resonance
2003 – 2006
Early career in software development and IT consulting, building the technical foundation across multiple domains and technologies.
My Projects
Here are some of my recent works
Autonomous Tool Discovery System
M.Tech dissertation exploring autonomous tool discovery through reinforcement learning and the Model Context Protocol. Achieved 50.33% task completion rate with 7.5% improvement over baseline strategies using Q-learning and DQN.
Why it matters: Demonstrates how AI agents can autonomously discover and compose tools -- a core capability for enterprise multi-agent systems.
RAG Financial Chatbot
Retrieval-Augmented Generation chatbot for answering financial questions based on company statements. Implements advanced RAG techniques including chunk merging, adaptive retrieval, BM25 search, and guardrails with a Streamlit web interface.
Why it matters: Shows how RAG enables domain-specific AI assistants that enterprises can trust -- grounded in real data with built-in guardrails.
MLOps Pipeline with DVC
MLOps project implementing data version control for house price prediction models. Uses DVC for versioning datasets and models with Azure Blob Storage integration, ensuring reproducibility and collaboration in ML workflows.
Why it matters: Production ML requires reproducibility and governance -- this pipeline ensures models are versioned, traceable, and auditable.
Fashion MNIST MLOps Pipeline
Comprehensive MLOps workflow for Fashion MNIST featuring automated EDA, HOG feature engineering, model explainability with LIME/SHAP, hyperparameter optimization using Optuna, and production monitoring with MLflow for drift detection.
Why it matters: End-to-end MLOps with explainability and drift detection -- the kind of production-grade pipeline enterprises need for responsible AI.
Autoencoder Variants for Feature Extraction
Comparative study of dimensionality reduction techniques — Standard/Randomized PCA vs Tied-Weight and Deep Convolutional Autoencoders on CIFAR-10 and MNIST. Achieved 0.989 cosine similarity validating PCA-autoencoder equivalence, and orders-of-magnitude reconstruction improvement with deep convolutional architectures.
Why it matters: Understanding when deep learning outperforms classical methods is key to choosing the right tool for enterprise feature engineering.
Advanced GAN Implementations on CIFAR-10
Systematic comparison of four GAN architectures — WGAN-GP, SNGAN, SAGAN, and SAGAN without spectral normalization (ablation study). SNGAN achieved best FID of 35.1; ablation study demonstrated spectral normalization's critical role in preventing mode collapse.
Why it matters: Rigorous ablation studies like this build the empirical foundation for selecting generative AI architectures in production settings.
Dice Game — Reinforcement Learning
Solved a stochastic dice game using Dynamic Programming. Implemented Policy Iteration and Value Iteration to find the optimal strategy (roll at 1–19, stop at 20–99) in a Markov Decision Process, validated over 10,000 simulations.
Why it matters: RL fundamentals like MDPs and dynamic programming underpin the decision-making logic in modern AI agents and autonomous systems.
Skills & Technologies
The toolkit I bring to AI transformation
AI & Emerging Technologies
ML Engineering & MLOps
Enterprise Transformation
Agile & Program Management
Domain Expertise
Certifications & Education
Continuous learning across AI, Agile, and cloud
AI & Generative AI
- Google Cloud -- Generative AI Fundamentals
- Google Cloud -- Introduction to Large Language Models
- Google Cloud -- Attention Mechanism
- Google Cloud -- Introduction to Image Generation
- Google Cloud -- Introduction to Responsible AI
Agile & Transformation
- Certified Scrum Professional (CSP)
- Certified ScrumMaster (CSM)
- Certified Scrum Product Owner (CSPO)
- Scrum@Scale Practitioner
- SAFe 4 Practitioner
- OKRs -- Objectives & Key Results
Cloud & DevOps
- DevOps Foundation
- Cloud Computing Foundation
- ITIL Foundation
Education
- M.Tech in AI & Machine Learning -- BITS Pilani
- MCA -- IGNOU
- PG Diploma in Advanced Computing -- C-DAC
- BCA -- MCRP University
Let's Talk AI Transformation
Whether you're starting your organization's AI journey, scaling AI adoption, or exploring multi-agent systems -- I'd love to connect.
Let's Connect
I'm always interested in conversations about enterprise AI strategy, multi-agent systems, and helping organizations build real AI capabilities. Whether you have a transformation challenge or want to explore what's possible with agentic AI, feel free to reach out!
Location
Pune, India