19.10.2026 aktualisiert


Premiumkunde
20 % teilweise verfügbarFull Stack AI Engineer | Python | AWS (2x Certs) | Terraform | React | TypeScript
Berlin, Deutschland
Berlin +25 km
BScÜber mich
With 12+ years of software engineering experience and a background in full-stack and cloud development, I now focus on agentic systems. I work primarily in Python and React, deploy on AWS and GCP, and lean heavily on agentic development workflows (Claude Code) day-to-day.
Skills
HTMLAPIsKünstliche IntelligenzAmazon Web ServicesCloudfrontAmazon S3Bash ShellCloud-EngineeringCode-ReviewDatenbankenContinuous IntegrationDevOpsGitHubInfrastrukturPython
Provided contact and working languages: English, German. Ca. 10 - 20% availability until Dec 2026
Primary Focus Areas
AI / LLM Engineering Building LLM-powered applications — tool-calling agents, agent loop architecture, MCP integrations, and RAG. Production concerns: SSE streaming APIs, multi-turn context management, and LLM observability (OTel-compatible traces, metrics, latency/error tracking). Daily agentic dev workflows with Claude Code.
Full-Stack/Backend Development
12+ years of hands-on experience - Led full-stack development across multiple technology stacks, from early-stage MVPs to applications within enterprises.
DevOps
5+ years implementing CI/CD and infrastructure automation - Designed and maintained deployment pipelines using GitHub Actions since 2020, implemented infrastructure-as-code with Terraform, and established automated testing practices across teams. Experience with Kubernetes orchestration and container management.
Cloud Engineering
AWS Solutions Architect Associate certified - Deep hands-on experience with AWS services including ECS, Aurora, VPC design, and serverless architectures. Managed multi-environment deployments and implemented security best practices. Growing expertise with Google Cloud through ongoing side-projects. Experience with Azure through previous projects.
Data Engineering
Data engineering spanning traditional and emerging approaches - Proven experience with database design, basic ETL pipelines (pandas/SQL), and statistical analysis.
Core Technical Competencies
- AI / LLM: Claude API, Hugging Face; agents & tool-calling, MCP, RAG, LangChain/LangGraph, structured outputs; SSE streaming; LLM observability (LangSmith, OTel)
- Web Frontend: React.js, Vue.js, TypeScript, JavaScript, Material UI - 7+ years production experience
- Backend: Python (FastAPI, SQLAlchemy, Pydantic), Node.js, previous statically-typed OOP experience (C#, both dotnet core and pre-core)
- API Design: RESTful services, authentication (Auth0, Firebase)
- Cloud & Infrastructure: AWS, ECS, Aurora, S3, CloudFront, VPC, CloudWatch, Secrets Manager
- Infrastructure as Code: Terraform (HCL), 2+ years production usage
- Containerization: Docker, Kubernetes (deployments)
- DevOps & Automation, CI/CD: GitHub Actions, GitLab CI, automated testing pipelines
- Testing: End-to-end (Playwright, Cypress), unit testing across multiple frameworks
- Monitoring and Observability: Application observability, infrastructure health checks, recent exposure to DataDog
- Data & Databases: PostgreSQL, MySQL, MongoDB, Supabase, ORMs, schema design and optimization
- Data Processing: pandas, SQL scripting, data pipeline development
- Visualization: matplotlib, d3.js, nivo
- Leadership & Process: Technical mentoring, code review processes, agile methodologies
- Architecture planning, technology evaluation, knowledge transfer
Erfahrung
Berufserfahrungmehr als 10 Jahre
Freelancer seit2025
Sprachen
DeutschverhandlungssicherEnglischMutterspracheJapanischGrundkenntnisseSchwedischGrundkenntnisse
Projekthistorie
Summary: A CLI tool supporting declarative .env management for worktree and branch-based workflows (human or agentic).
Repository: https://github.com/AndrewADev/bight
Key Features & Accomplishments
- Develop user-friendly flows including atomic temp-file-plus-rename writes, Unix permission and comments preservation, as well as optional backups.
- Implement a declarative configuration schema supporting both granular project-local settings and global defaults.
- Automate release pipeline via GitHub Actions to publish pre-compiled binaries and verifications across macOS, Linux, and Windows, automatically updating a companion Homebrew tap on new tags.
- Integrate cobra for CLI structure, including a bight install subcommand that registers the post-checkout git hook automatically.
Technologies
- Backend: Go
- DevOps: GitHub Actions
- Other Tools: cobra, godotenv, yaml.v3, Homebrew
Duration: Feb 2025 - Present
Effort: 60 - 80d
Key Features & Accomplishments
- Develop full-stack learning application using React with Material UI and TypeScript frontend, implementing Redux Toolkit for state management and Firebase Authentication for user management.
- Implement scalable cloud infrastructure on Google Cloud Platform using Terraform IaC, featuring scalable containerized deployment with Cloud Run, automate image management via Artifact Registry, and secure CI/CD pipelines using GitHub Actions with workload identity.
- Build comprehensive RESTful API with FastAPI featuring dependency injection, robust error handling, and pagination support for CRUD operations with user-scoped data access.
- Establish modern development workflows including Docker containerization, multi-stage builds for production optimization, automated artifact cleanup policies, and infrastructure-as-code practices for reproducible deployments.
Technologies
- Frontend: React.js, TypeScript, Material UI, HTML, CSS, Zod, yarn
- Backend: Python, FastAPI, Pydantic, Firebase (Auth)
- Data: PostgreSQL, Alembic, SQLAlchemy, Google Cloud (Big Query)
- Infrastructure: Google Cloud (Cloud Run, Artifact Registry, GCE, Budgets), docker, Terraform, HCL, Traefik, Neon
- DevOps: uv, GitHub Actions, Task & Taskfiles, Renovate
- Testing: pytest, Playwright, Storybook
Effort: 15-20 d
Summary: Agentic flows built with LangChain and LangGraph that explore and analyze GitHub repositories using configurable LLM providers (HuggingFace/Ollama/Anthropic) for automated repository research and insights.
Key Features & Accomplishments
- Develop multi-provider LLM agent using LangChain and LangGraph with SQLite-based state persistence, supporting Ollama (local), Anthropic (cloud), and HuggingFace Inference API for flexible deployment and privacy options.
- Implemented comprehensive GitHub API integration featuring starred repository pulse analysis, topic-based search, commit hotspot detection with pluggable churn calculation strategies (total activity vs. rework rate), and thread-based conversation management with resume capability.
- Build dual-interface system with Typer CLI for scripting/automation and Gradio web UI for interactive exploration, including conversation history browser and repository favorites management..
Technologies
- Backend: Python, LangChain, LangGraph, Anthropic API, Pydantic, Typer, SQLite
- Infrastructure: Ollama, HuggingFace
- DevOps: uv, ruff, prek, pipenv, python-dotenv, LangSmith
- Testing: pytest, pytest-mock
- Other Tools: Claude Code