Asif Hussain

Forward Deployed AI Engineer

Upwork • Freelance

Jan 2026 - Present • 4 mos

Embedded engineer for product teams shipping LLMs and agentic AI into production. I sit between the engineering team and their AI stack — scoping use cases into shippable integrations, building the system, hardening it for real users, and handing off documented code the in-house team can own and extend. What I deliver: - Scoping & technical discovery — translating "we want AI in our product" goals into specific, measurable, shippable integrations with clear success metrics, evals, and acceptance criteria - Production-grade AI agents using LangChain and LangGraph — real tool calling, state management, human-in-the-loop checkpoints, proper retry and timeout handling, structured logging, and cost tracking baked in from day one - Custom MCP (Model Context Protocol) servers for Claude Desktop, Cursor, and IDE integrations — built in Python and TypeScript using the official Anthropic SDKs, with OAuth, rate limiting, audit logging, and full setup documentation - RAG (Retrieval Augmented Generation) systems with Pinecone, Weaviate, and pgvector — including chunking strategy, hybrid search (BM25 + vector), reranking, citation tracking, and evaluation harnesses to measure retrieval quality on real data - LLM cost optimization audits — typical engagements cut OpenAI/Anthropic API bills by 40-80% with zero quality regression. Methodology covers model right-sizing (Claude Haiku vs Sonnet, GPT-4o-mini vs GPT-4o), prompt caching, batch APIs, context window optimization, and embedding cache strategies - Multi-agent orchestration with LangGraph — planner/researcher/ synthesizer patterns, parallel tool execution, state machine design, and recovery from partial failures - Streaming UX implementation — server-sent events and WebSocket patterns for real-time token streaming, partial tool calls, and responsive AI interfaces - Production debugging for AI systems — diagnosing hallucinations, JSON parse failures, broken

Full Stack Developer

Finance Expert • Full-time

Dec 2022 - Present • 3 yrs 5 mos

Built and shipped production web applications for clients across SaaS, e-commerce, fintech, and marketplace domains. End-to-end delivery from architecture through deployment, with a focus on payment integrations, REST/GraphQL API design, and scalable backend systems. Core work: - Full-stack web application development using the MERN stack (MongoDB, Express, React, Node.js) — built dashboards, admin panels, customer portals, and B2B SaaS products from scratch - Backend API development with Node.js (Sails.js, Express) and PHP (Laravel) — designed RESTful and GraphQL APIs serving mobile, web, and third-party integration clients - Payment integration specialist with Stripe — subscriptions, one-time payments, Stripe Connect for marketplaces, webhook handling, refund flows, dispute management, PCI-compliant tokenization, and SCA/3D Secure compliance - Frontend development with React, Next.js, and Angular — state management (Redux, Context API), responsive design, form handling, and integration with backend APIs - Database design and optimization — PostgreSQL, MySQL, MongoDB. Schema design, indexing strategy, query optimization, and migration management - Third-party API integrations — Stripe, PayPal, Twilio, SendGrid, AWS S3, Google Maps, OAuth providers, and many domain-specific APIs depending on client need - Authentication & authorization — JWT, OAuth2, session-based auth, role-based access control, and multi-tenant architecture patterns - DevOps and deployment — Docker containerization, CI/CD with GitHub Actions, deployment to AWS (EC2, ECS, Lambda), Heroku, DigitalOcean, and Vercel - Code quality practices — unit and integration tests (Jest, Mocha, PHPUnit), code reviews, linting (ESLint, Prettier), and Git workflows with feature branching Stack: Frontend: React · Next.js · Angular · TypeScript · JavaScript · HTML5 · CSS3 · Tailwind Backend: Node.js · Express · Sails.js · Laravel · Python