Experience

• Designed and implemented a risk analysis system using ML (scikit-learn, XGBoost) and agentic AI.
• Architected parallelized AI/Agentic workflows achieving ~3x faster execution with higher reliability.
• Automated the end-to-end KYC lifecycle improving onboarding, fraud detection, and document verification.
• Reduced OCR provider costs by up to 30x via multimodal LLM integrations; improved accuracy and throughput.
• Improved document processing speed and resilience with parallelism and exponential backoff retries.
• Delivered a voice/text customer support system (Django, Gemini, Twilio, S3 vectors) with React/Vite/shadcn UI.
• Deployed on AWS ECS with Docker and Terraform; automated via GitHub Actions and instrumented with Grafana.

• Built a static site with Astro and deployed to Cloudflare Pages.
• Enforced performance budgets with Lighthouse CI.

Company

• Developed backend services with Django and frontend components with React.
• Implemented AI-driven PoCs focusing on NLP applications.

• Built generative AI applications with Django and LangChain; React/TypeScript frontend with Tailwind.
• Added agentic capabilities to an enterprise chatbot for tabular data analysis (LangGraph, tool calling).
• Developed blog generation and competitor analysis tool for a publishing firm.

• Built an AI-evaluated multiplayer game ('Death by AI') using PlayroomKit SDK and React.
• Improved accessibility for a legacy Angular 1.4 frontend.
• Implemented RAG pipelines: FAISS for small-scale, Milvus for large-scale vector searches.
• Migrated authorization from RBAC to ABAC with CASL across React and NestJS services.
• Implemented an end-to-end change request ticketing system (React + Nest).