Work
8+ years across AI research, applied ML, and enterprise GenAI architecture in financial services.
Experience
Gen AI Architect · SBER Bank India
- Led the team building GigaOCR, a multimodal data extraction system (Hugging Face Transformers, olmOCR, vLLM, Qwen3-VL) for scanned documents, images, and handwritten invoices.
- Contributed to GigaChat enterprise LLM solutions — role-based chatbots for HR, IT, Treasury, Methodology, Credit, Investment, and Sales.
- Implemented multi-turn dialogue context modeling, state tracking, and a voice-agent framework with speech-to-text interaction.
Senior AI Engineer · Ernst & Young LLP
- Built a multimodal AI compliance platform for a global chemicals client, analyzing product regulations across 50+ countries using CLIP, FAISS, Neo4j knowledge graphs, and GPT-4.
- Ran supervised fine-tuning and alignment of domain-specific LLMs; implemented agentic workflows with memory, tool-calling, human-feedback loops, and recovery mechanisms.
- Achieved roughly 70x faster compliance processing versus the manual baseline.
Senior Data Scientist · Société Générale
- Built an AI-driven onboarding and compliance verification system using Tesseract OCR and LLMs for document parsing, risk profiling, and KYC summarization, deployed on AWS.
- Developed a production predictive model for justifying Net Asset Value (NAV) in investment management using decision trees, RabbitMQ, PostgreSQL, Airflow, Docker, and Kubernetes.
Applied Machine Learning Engineer · TCS Research
- Built an NLP-based query chatbot (Django) and a next-word-prediction module using an LSTM sequential model (Keras/TensorFlow/PyTorch).
- Handled large-scale ML training and prediction on Hadoop/Spark distributed systems, with A/B testing and MLOps for lifecycle management.
Applied Machine Learning Engineer · GateOverflow
- Ran semantic analysis on user comments with Python and NLTK to extract sentiment, intent, and key themes for customer feedback insights.
Skills
AI/ML: CNN, RNN, LSTM, XGBoost, AutoML, Predictive Modeling, Time Series
GenAI: Transformers, Prompt/Context Engineering, RAG, Agentic AI, SFT, FFT, PEFT, LoRA/QLoRA, RL Reward Optimization, MCP, OCR, vLLM
Programming: Python, SQL, JavaScript
Frameworks: TensorFlow, PyTorch, Flask, Django, Streamlit, FastAPI
DevOps: Git, Docker, Kubernetes, Airflow, CI/CD, MLflow, DVC
Cloud & Data: AWS, PostgreSQL, Spark, Hadoop, Azure, GCP
Education & certifications
B.Tech, Information Technology — Kamla Nehru Institute of Technology (KNIT), Sultanpur, IN (2014–2018).
Certified Machine Learning Expert (AppliedAI) · GATE Qualified, 97.10 percentile · Certified Deep Reinforcement Learning Expert (Udemy) · Speaker, Open Machine Learning Conference, Bengaluru.
Products & consulting
I take on a limited number of advisory and delivery engagements outside my full-time role — architecture reviews for RAG/agentic systems, LLM evaluation and fine-tuning strategy, and AI-vendor or AI-talent sourcing for teams building out a GenAI function.
Get in touch