Presentation
## What I offer:
Full-stack web development with Next.js, React, TypeScript, and Tailwind CSS
Backend API development with Python (FastAPI, Django), Node.js, and Express
AI/LLM integration including LangChain, LangGraph, RAG pipelines, and structured output extraction
Cloud deployment and infrastructure on GCP, Azure, Docker, and CI/CD pipelines
Mobile development with React Native
Recent work includes:
Medical document processing pipeline that extracts structured data from 15+ page PDFs using OCR and AI (Pydantic, Google Gemini, Django Ninja, Celery, GCP)
AI concierge platform with 7+ microservices handling hotels, flights, calendars, and restaurants, cutting LLM costs by 67% (LangGraph, MCP, Zod, TypeScript)
Developer talent platform serving 800+ users with profile management, resume parsing, and AI-powered profile rewriting (Next.js, TypeScript, GCP)
Tech stack: TypeScript, Python, JavaScript, Next.js, React, FastAPI, Django, Node.js, LangChain, LangGraph, PostgreSQL, Redis, Docker, GCP, Azure
Portfolio
Background
- AOSN: Managed Render infrastructure and Sentry diagnostics; re-engineered Stytch MFA auth flows and resolved backend session errors to eliminate 401 regressions.
- PrivyAI: Optimized AI agent provisioning and RunPod/Docker runtimes; integrated Web Speech API and privacy-safe observability for real-time agent interactions.
- JanaIntake: Built medical document pipelines using Django Ninja and Celery; utilized Gemini and Pydantic to automate clinical entity extraction and ICD-10 mapping.
- Tech stack: TypeScript, JavaScript, Python, Next.js, React, FastAPI, Sanity CMS, PostgreSQL, Docker, RunPod, Stytch, Sentry, Plausible, PostHog.
- Architected a modular microservice layer using the Model Context Protocol (MCP) to integrate 10+ third-party APIs (Google Maps, Ticketmaster, Duffel, RateHawk), enabling AI agents to execute complex, multi-step booking workflows.
- Streamlined agentic state management by implementing LangGraph, replacing rigid top-down hierarchies with a Swarm architecture to facilitate dynamic task handoffs and autonomous inter-agent communication.
- Enhanced retrieval accuracy by integrating RAG pipelines with vector search for semantic queries across events and todos, successfully replacing legacy keyword-based retrieval systems.
- Developed a high-concurrency browser automation service using FastAPI and browser-use to manage parallel Chrome instances and profile persistence for reservations across Google Maps, OpenTable, and SevenRooms.
- Tech Stack: LangGraph, LangChain, Model Context Protocol (MCP), Pinecone, RAG, Express.js, Laravel, TypeScript, JavaScript, Cosmos DB, Azure, Firebase, Playwright,
Python, FastAPI, browser-use
- Led product direction for eskalate.io, planning roadmaps and executing Agile sprints based on 90+ developer interviews.
- Boosted profile completion from 13% to 90%+ by adding contextual tooltips, simplifying editors for consistent formatting, and conducting targeted outreach.
- Rebuilt the marketing landing page with service sections, navbar dropdowns, and responsive service pages.
- Dockerized the application with multi-stage builds for GCP deployment.
- Tech Stack: Next.js, TypeScript, Tailwind CSS, Shadcn UI, Docker, Jira, GCP
Frontend Developer · Oct 2023 – Jul 2024
- Architected the frontend from scratch — folder structure, Redux/RTK Query state layer, refresh token system, coding guidelines, linters, CI/CD, and GitHub configurations.
- Built the complete profile system (20+ components): multi-step forms with validation, drag-and-drop resume upload, image cropping, and progress tracking for 800+ users.
- Led frontend contributions across a 5-developer frontend team within a 9-developer cross-functional team.
- Tech Stack: Next.js, TypeScript, Redux Toolkit, RTK Query, Tailwind CSS, GCP
• Solved 600+ problems on LeetCode and Codeforces, strengthening data structures and algorithm expertise.


