Martin H.

Django and AI engineer, Top Rated Plus with 2,600+ hours and 100% Job Success on Upwork. I add production AI (RAG, LLM apps, automation) to existing Django systems, then keep them running in production If you have a live product and need someone who can ship a new AI feature without breaking what already works, or who can own the backend end to end from architecture through deployment and daily operations, that is the work I do best. Recent work: - Industrial IoT control interface, live in production for over a year, with 1,200+ hours on one ongoing contract. An online control interface for hardware, built to stay reliable under continuous use. - A RAG memory system built on real document retrieval and embeddings. It is in my portfolio as working software, not a demo slide. - A support desk AI that transcribes and summarizes client phone calls and drafts the reply for a human to send, wired into voice and SMS. - Promoted mid-contract from assistant to full stack developer after consistently shipping difficult work. The client's own words: reliable, talented, hard-working, excellent communication. What I build: - Backend and API: Django, Flask, REST APIs, PostgreSQL design and optimization, Redis, Celery for async and scheduled work. - AI and automation: RAG and embeddings, LLM APIs (OpenAI, Claude, Gemini), call transcription and summarization, chatbots, and workflow automation that removes manual steps. - Infrastructure and DevOps: AWS (EC2, S3, Lambda, ECR, CloudWatch), Docker, CI/CD with GitHub Actions and Ansible, and monitoring with Grafana so problems surface before users report them. - Integrations: Twilio and BulkVS for voice and SMS, QuickBooks Desktop through Conductor, and third-party APIs where the data lives. How I approach a project: I start by understanding what you are trying to ship and what is already running, because the risk on a live system is usually the code you did not touch. I keep communication steady, flag blockers early, and give honest estimates with the caveats stated plainly. I would rather set a conservative date and beat it than promise fast and slip. When something falls outside the original scope, I say so and we agree on it before I build, so there are no invoice surprises later. Who I am the best fit for: Startups and growing teams that want one engineer to own their Django backend and its AI features for the long haul, instead of handing a critical system to someone new every few months. Questions clients usually ask me: How soon can you start, and what does the first week look like? Usually within a few days. I start by reading the codebase and the one or two things that are actually on fire, then send you a short plan before I change anything, so you see the scope and the risks before the meter really runs. My codebase is messy and only half documented. Can you still work in it? Yes, that is most of what I do. Adding a feature to a live system I did not write is the normal case, not the exception. I read before I touch, keep changes small and reversible, and tell you when something is risky instead of quietly working around it. Do you work solo, or with our existing team? Both. I can own a backend end to end on my own, or plug into your team's repo, standups, and review process and match how you already work. I am comfortable taking direction from a lead or setting the technical direction myself. Do you do the work yourself? Yes, I write the code. I use AI tools like Claude Code and Codex to move faster on the routine parts, but the architecture, the decisions, and the responsibility are mine. You are hiring one engineer who stays with your project, not a rotating cast. Core stack: Python, Django, Flask, PostgreSQL, Redis, Docker, AWS, GitHub Actions, Ansible, Linux. AI: RAG, embeddings, LLM APIs, transcription, agent and automation workflows. Tell me your stack and what you are trying to ship, and I will tell you honestly whether I am the right fit.