Munich, Germany

AnasHabib —builds AI systems

Software engineer with 2+ years building full-stack products and 1+ year shipping production AI. React and Next.js frontends, TypeScript and Python APIs, and the agents, retrieval and evaluation behind them, on AWS.

Years engineering
2+
Saved / week
10h
Contracts indexed
100+

What I build

01 — Retrieval

Finding the right context

Documents become vectors. A question becomes a vector too. The neighbourhood that lights up is what the model actually gets to see — get this wrong and nothing downstream matters.

  • pgvector
  • Pinecone
  • LangChain
  • RAG

02 — Extraction

Messy in, schema out

Scanned receipts flow through extract, validate, store. What the model isn't sure about lands in a review screen for a person instead of being quietly guessed at.

  • Multimodal LLM
  • JSON Schema
  • NestJS
  • React

03 — Agents

Tools, not chat

A router reads the question and hands it to one of three specialist agents. Agents propose changes; an authorized person confirms each one before it is saved.

  • OpenAI Responses API
  • Tool calling
  • Human-in-the-loop
  • NestJS

04 — Shipped

Running in production

Three production apps at ProSiebenSat.1, an AI consulting platform with paying customers, and a multi-tenant workspace backend. All on AWS, shipped through automated CI/CD.

  • ECS Fargate
  • Docker
  • Terraform
  • GitLab CI/CD

Where I have shipped

Experience

  1. Software Engineer

    HybridRedseven Entertainment GmbH(ProSiebenSat.1 Group)

    Munich, Germany · Mar 2026 — Present

    • Led full-stack development of three production apps using TypeScript, NestJS, Next.js, React and PostgreSQL on AWS, including a licensing platform for 30+ staff launched in three months
    • Built an AI assistant for rights and production teams that answers licensing cost questions in one step instead of a 30–60-minute manual search, using NestJS, the OpenAI Responses API and 3 specialist agents
    • Shipped a receipt-processing app for 50+ finance staff in React and NestJS, saving 10+ hours per week
    • Engineered an AI extraction pipeline that reads receipt images with a multimodal LLM into JSON-schema structured outputs, reaching 90% field-level accuracy across 20 country tax codes
    • Designed an agentic RAG pipeline on pgvector over 100+ contracts and policies, so every answer cites its source
    • Implemented an LLM evaluation pipeline in Braintrust that blocks any release below 70% faithfulness and context recall (scoring 85%), catching a routing bug before launch
    • Automated deployments with GitLab CI/CD, Docker and automated tests on AWS ECS Fargate, cutting deployment time by an estimated 70% and eliminating failed deploys
    • Developed a human-in-the-loop approval flow in React with RBAC, where authorized users confirm every AI-proposed change before it is saved, preventing an incorrect licensing update

  2. Software Engineer

    HybridArcpeak

    Munich, Germany · Aug 2025 — Feb 2026

    • Led full-stack development of an AI consulting platform using Python, FastAPI, React, TypeScript and PostgreSQL, turning 20 client answers into 10 ranked AI opportunities
    • Integrated Stripe subscriptions and OAuth sign-in into the FastAPI platform, enabling its first paying customers
    • Developed a Python AI agent with live web search that turns each opportunity into a cited adoption plan
    • Architected a background job system with Redis Streams that keeps long AI reports running after users close the browser, streaming each section to the React dashboard
    • Automated agent release checks in CI with LLM evaluation and Pytest tool-call tests, stopping faulty releases

  3. Software Engineer

    RemoteWorkSpin

    Barnet, UK · Jun 2024 — Jul 2025

    • Led backend development of Boarddd, a multi-tenant business workspace, using Node.js, Express, MongoDB, Redis and Socket.IO on AWS, with frontend and QA engineers
    • Optimized MongoDB schemas, indexes and Redis caching, cutting query latency by 70% and peak load by 50%
    • Built a Node.js payments microservice behind service-to-service authentication on 3 Stripe products (Connect, Treasury, Issuing) for subscriptions, recurring invoices and virtual cards
    • Engineered integrations with 8 external tool APIs (Jira, Asana, Trello and more) on a fault-tolerant BullMQ job pipeline with rate limiting, retries and live progress
    • Designed a scalable real-time messaging backend with Socket.IO and MongoDB for group channels of 50+ members and direct messages, with threaded replies, reactions and paginated queries

Built on my own time

Projects

InsightQL

AI Database Assistant

A natural-language database assistant that lets non-technical users query three database engines (PostgreSQL, SQLite, MongoDB) in plain English through a LangChain SQL agent, with charts, CSV/Excel export and JWT-secured access.

  • Next.js
  • NestJS
  • LangChain
  • OpenAI
  • Docker

bugSage

AI Debugging Assistant

An AI debugging chatbot that retrieves relevant docs and past issues from a vector database before answering, so its Express.js fixes match the project's own code.

  • FastAPI
  • Pinecone
  • RAG

CLI Assistant

Agentic Terminal Tool

A fully offline terminal assistant on a local model, with an agentic loop that chains five tools to complete multi-step tasks.

  • Python
  • Ollama
  • Function calling

The toolset

Skills

Programming Languages
PythonTypeScriptJavaScriptSQL
Backend & Frameworks
Node.jsNestJSFastAPIExpressREST APIsSocket.IOBullMQ
Frontend
ReactNext.jsRedux ToolkitTailwind CSS
Databases
PostgreSQLMongoDBRedispgvector
AI/ML
AI AgentsAgentic RAGLLM EvaluationFunction CallingOpenAI APILangChainLangGraph
Cloud & DevOps
AWSDockerTerraformGitLab CI/CD
AI Tools
Claude CodeCursorGitHub CopilotCodex

Education

Studied

  • University of Passau

    MSc in Computer Science

    Passau, Germany · Oct 2024 — Oct 2026

  • FAST-NUCES

    BS in Software Engineering

    Karachi, Pakistan · Aug 2020 — Jun 2024

Published

Research

Ris-Enhanced 6G Networks: Advanced User Location Prediction with Generative Models

2024 IEEE 9th International Conference on Engineering Technologies and Applied Sciences (ICETAS)

Compared GRU, LSTM and Transformer models for user-location prediction in RIS-assisted 6G networks; the Transformer achieved the lowest mean absolute error in the study.

Open to software engineering and applied AI roles

Let’s talk

Download CV (PDF)