Enterprise AI leader with 19+ years building production-grade GenAI, Agentic AI, Computer Vision, and NLP systems. Executive AI programmes delivered at Deloitte India, Ernst & Young, and UAE Government (ADAA). Teaching graduate AI at a US university while advising Fortune 500 FMCG companies on AI strategy.
I stand at the intersection of three disciplines that rarely converge in one career — I build and ship production AI systems, teach it at graduate level in the US and India, and advise Fortune 500 FMCG companies on AI-driven strategy. This combination is genuinely rare at the leadership level.
“Most AI experts advise on what I have actually architected, deployed, and measured in production.”
Over 19+ years I have moved from analytics foundations at Genpact through strategic leadership at Great Learning, where I run AI/ML and GenAI product delivery across four cities while simultaneously teaching Computer Vision at Northwestern University (MSDS) and NLP & Generative AI at Great Lakes Institute of Management.
I bring equal fluency in transformer architecture, CXO-level strategy conversations, and classroom delivery — applying the same depth whether presenting an AI portfolio to a CMO or explaining RAG internals to a senior engineer.
Ships production RAG pipelines, Agentic AI systems, and LLM-powered products with measurable outcomes. 30% SLA improvement. Predictive ML at scale.
Adjunct Professor at Northwestern University (Computer Vision, MSDS) and Great Lakes Institute of Management (NLP & GenAI). 5-star mentoring rating.
Delivered executive GenAI at Deloitte India, UAE Government, and Ernst & Young. 3,000+ professionals trained across India and internationally.
Designed 15-product AI portfolios and I2M roadmaps for Fortune 500 FMCG clients with sequenced CXO pitch structures and PoC-first deployment blueprints.
A 19-year arc from analytics foundations to AI product leadership — each role building the next layer of depth.
Built across 19+ years of applied research, enterprise delivery, and graduate-level teaching — not certifications.
End-to-end LLM strategy — model selection, RAG pipeline architecture, fine-tuning, and production deployment. Shipped LLM-powered systems reducing SLA by 30%. Delivered GenAI strategy to Fortune 500 teams and Big-4 leaders. Teach LLM selection as a practitioner, not a theorist.
Designed and delivered Agentic AI programmes for Deloitte India and the UAE Government. Currently delivering at Ernst & Young. Built multi-agent supply chain blueprints with real enterprise deployment structures and autonomous agent orchestration patterns.
Teaching CV at Northwestern University MSDS (DSP 462) — Vision Transformers, object detection, segmentation, ONNX edge deployment — while building real-world CV systems for infrastructure defect detection. Full pipeline from model training to edge inference.
Deep expertise from n-gram statistical models through Word2Vec, BERT, and GPT-scale architectures. Teaching NLP & Generative AI at Great Lakes for executive cohorts. Published practitioner-grade articles on NLP evolution and transformer mechanics used in corporate training programmes.
My academic work extends my industry practice — every concept taught has been applied in production environments.
Teaching DSP 462 — Computer Vision to the MSDS graduate cohort. Vision Transformers, transfer learning, ONNX edge deployment, Faster R-CNN, and LLM-vision integration. Assessments designed around production deployment readiness.
Delivering NLP & Generative AI to working professionals and executive cohorts. Transformer architecture, BERT, RAG pipeline design, LLM fine-tuning, and enterprise GenAI deployment strategy structured for immediate applicability.
Applied AI & Deep Learning sessions for engineering students — CNN architectures, transfer learning, and production deployment patterns bridging academic theory with industry practice.
Every programme built from scratch for the client — calibrated to the exact audience, anchored in real business decisions, not generic AI slides.
Delivered to a mixed audience of cybersecurity professionals, business leaders, IT managers, and operations heads. Problem-first methodology — real business scenarios anchored every module. Covered enterprise RAG deployment, prompt security, LLM selection, and Agentic AI workflow design.
Delivered to Accounting, Audit, Cybersecurity, and Business Leadership teams planning government-scale AI deployment. Covered autonomous agent design, tool orchestration, multi-agent pipelines, and Sovereign AI architecture for national-level implementation.
Confirmed enterprise Agentic AI programme for Ernst & Young. This engagement positions Mukul as one of the very few practitioner-educators trusted by multiple Big-4 consulting firms for AI capability building at the leadership level.
56 public repositories on GitHub. Four selected highlights spanning LLM applications, Computer Vision, and enterprise AI resources.
LLM-powered resume–JD semantic alignment tool using Hugging Face transformers. Projected +30% improvement in ATS shortlisting rates and +25% application success. A live demonstration of RAG + semantic search in a high-stakes real-world use case.
View on GitHub →Deep learning CV system for AI-based crack detection in concrete infrastructure — applying CNN architectures and transfer learning to structural safety classification. Demonstrates applied Computer Vision in a real-world engineering and safety context.
View on GitHub →Deployed enterprise AI reference micro-site covering LLMs, RAG pipeline design, vector embeddings, prompt engineering, and responsible AI. Used directly as live training material in both Deloitte India and ADAA UAE programmes.
Visit the Site →Multimodal AI pipeline transforming images into structured insights using Google Gemini. Demonstrates tool-use, multimodal reasoning, and API integration patterns — used as live teaching material for GenAI and multimodal AI modules across Great Learning and corporate training programmes.
View on GitHub →Writing that gives a data scientist technical depth and a business leader strategic clarity — in the same article.
A decision framework every enterprise team should read before choosing between LLMs and traditional ML. Covers cost, latency, task fit, and governance.
Read on Medium →Traces the full arc of Natural Language Processing — from statistical n-gram models through Word2Vec to transformer-era breakthroughs.
Read on Medium →The architectural revolution that changed how machines understand language — attention mechanisms, bidirectional training, and transfer learning explained precisely.
Read on Medium →How vector databases underpin modern RAG systems — practical implementation patterns, indexing strategies, and tool comparisons for enterprise teams.
Read on Medium →Unpacking the mechanics behind language models from the simplest probability-based approaches to GPT-scale autoregressive generation.
Read on Medium →BERT attention, gradient boosting, feature engineering, statistical inference, and applied ML.
Browse All →Actively exploring Director of AI, VP of Artificial Intelligence, and Head of AI roles at MNCs where AI is a core business function. Open to India, Overseas and global remote.
Also available for enterprise AI training, keynote engagements, and AI strategy consulting on a selective basis.