14.05.2026 aktualisiert


AI Product Engineer | AI Agents • AWS • Node.js • TypeScript • Python • Go
Über mich
AI Engineer, 8+ yrs full-stack experience. Production code in TypeScript, Python, Go. Builds AI agent systems, voice pipelines, RAG apps, and serverless architectures. Cut latency from 600ms to under 50ms, reduced cloud costs 40%+. Custom AI tooling: Claude Code, MCP servers, AI agents.
Skills
AI/ML Engineering: AI Agents, MCP, Voice AI, LLMs, VLMs, LangChain, LangGraph, RAG Systems, MLOps, MLFlow, Knowledge Graphs, AWS Bedrock, AWS AgentCore, Pinecone, pgvector, ChromaDB, Neo4j, Embeddings, Reranking, Hybrid Search, OpenAI API, Anthropic Claude API, Hugging Face, n8n
AI-Assisted Development & Rapid Prototyping: Claude Code, Lovable, OpenAI Codex, v0, Cursor
Fullstack Development: TypeScript, Next.js, Nest.js, React, React Native, Python (FastAPI, Flask, Django), Go, Node.js, Express.js, GraphQL, REST, gRPC, Microservices, Serverless
Cloud Architecture: AWS (EC2, ECS, EKS, Fargate, Lambda, Step Functions, API Gateway, Bedrock, AgentCore, SageMaker, S3, DynamoDB, RDS, Aurora, ElastiCache, CloudFormation, CDK, CloudFront, Route 53, VPC, SQS, SNS, EventBridge, Kinesis, CloudWatch, IAM, Cognito, KMS)
Infrastructure & DevOps: Terraform, CDK, Pulumi, Docker, Kubernetes, Helm, ArgoCD, GitOps, GitHub Actions, GitLab CI/CD, CircleCI, Localstack, Supabase
Databases & Messaging: PostgreSQL, MySQL, MongoDB, DynamoDB, Redis, Elasticsearch, Neo4j, Apache Kafka, RabbitMQ, Amazon Kinesis, MQTT, Event-Driven Architectures
Security & Compliance: OAuth 2.0, OIDC, JWT, IAM, Cognito, RBAC, Multi-Tenant Security, Secrets Manager
Sprachen
Projekthistorie
Owned the AI orchestration layer: designed and implemented the multi-agent system coordinating LLM and VLM models for automated product analytics. Built the integration layer that fetched and normalised data from external systems into a format the agents could reason over, handling RRWEB session replay data at scale. Implemented the embedding and vectorisation pipeline, storing and indexing vectors with pgvector and using similarity search to surface relevant session context for each agent query. Responsible for agent routing logic, context assembly, and wiring MCP tool-calling so models could read and write structured output directly to downstream systems without human-in-the-loop steps.
Established an organisation-wide Claude Code setup, standardising AI-assisted development workflows across the engineering team and reducing time-to-production for new features significantly.
Technologies: LLMs, VLMs, AI Agents, MCP, LangGraph, LangChain, RAG, Pinecone, pgvector, Python, FastAPI, TypeScript, Next.js, React, AWS (Lambda, S3, DynamoDB), Docker, GitHub Actions, Claude Code
