Banking Domain Expert · AI/ML Engineer · Exploring AI Safety in Finance
- 🏦 8+ years inside core banking — Functional Architect at Fiserv, Senior BA at HSBC (UK), built teller platforms, compliance modules (CTR, BSA, KYC), and 25+ banking APIs
- 🎓 MBA (Business Analytics, Hult SF) · B.E. (Computer Science, University of Pune)
- 🤖 Building AI/ML tools on the side — local LLMs, RAG pipelines, deep learning, NLP classifiers
- 🔍 Interested in AI safety for financial systems — how LLMs fail in banking, compliance risks, evaluation gaps
- ✍️ Writing about the intersection of banking, AI, and safety
| Project | Description |
|---|---|
| Local-LLM-Arena | 🤖 Benchmark and compare local LLMs side-by-side |
| Local-TTS-Studio | 🎙️ Local text-to-speech studio with voice synthesis |
| PDFQuery-VectorDB | 📄 RAG-based PDF question-answering with vector DB |
| CelebSearch-langchain-OpenAI | 🔗 LangChain + OpenAI entity search |
| Movie-Recommender-System | 🎬 ML/NLP-powered movie recommendation engine |
| DeepLearning-Basics | 🧠 Foundational deep learning models & notebooks |
| SMS-spam-classifier | 📱 SMS spam detection with classic ML/NLP |
| Banknote-predictor-api | 💳 ML-powered REST API for banknote authentication |
| ML-algorithms | 📊 Core ML algorithm implementations in Python |
Why banking × AI matters to me:
- Architected a microservices teller platform across 2 core banking systems — saw firsthand how AI tools fail on compliance-critical workflows (CTR, BSA, KYC)
- Designed and validated 25+ REST API contracts at HSBC UK and Fiserv — understand the surface area where LLMs interact with real banking data
- Led functional design for compliance-critical modules where a wrong AI output isn't just a bug — it's a regulatory violation
- 7+ years of seeing the gap between what AI promises in banking demos and what actually survives production
Making AI in banking safe, auditable, and trustworthy.