Senior Full-Stack Software Engineer · AI-Powered Products
Senior Full-Stack Software Engineer with 11+ years of experience designing and building scalable web applications, now focused on LLM-powered products in regulated environments. Based in Ingolstadt, Bavaria, Germany.
I work across the stack — from React/Next.js frontends to serverless backends on AWS and GenAI on Google Cloud — and care about shipping reliably: clean release processes, feature-flagged delivery, and UIs that hold up to compliance scrutiny. I enjoy mentoring developers and shaping engineering practices across teams.
Trenkwalder — building TalentAI, an LLM-based candidate screening service.
- Microservices & event-driven architecture: Designing and building TalentAI as microservices communicating asynchronously via AWS SQS and RabbitMQ, decoupling services for independent scaling and resilient processing
- EU AI Act–aware product design: Designed neutral, non-judgmental UI for AI-generated assessments — replaced a red/green traffic-light flag system with a four-tier attention scale (none / low / medium / high), triple-encoded via label, shape and tone for accessibility and fairness
- Microsoft Graph integration: Email integration delivered; building Outlook calendar integration (view availability, create interview events, handle candidate declines and reschedules) using app-only permissions
- Release engineering: Moved the team from a Gitflow-style model to trunk-based development with feature flags, tag-triggered deployments, immutable artifact promotion and a defined hotfix flow
- Feature flag architecture: Designed a two-tier PostHog flag setup — shared flags centralised, app-specific flags per project — evaluated via direct HTTP calls without an SDK
- Serverless config management: Strategies for propagating AWS SSM Parameter Store changes to Lambda (forced cold starts, EventBridge-driven invalidation, deploy-time baking)
- Monorepo CI/CD: Per-service tag conventions (e.g.
backend/v1.2.0) and pipeline scoping - Qualina voice agent: Building a real-time conversational voice agent on the Gemini Live audio API — designing guardrails to keep conversations on-scope and compliant, tool calling to connect the agent to backend actions, prompt engineering for consistent, natural dialogue, and tuning voice-activity detection to avoid self-interruption
Technologies: React, Next.js, Angular (7–18+), TypeScript, JavaScript, NgRx, Zustand, Apollo GraphQL, accessible & compliance-aware UI design
Technologies: Node.js, Java (Spring Boot), Kotlin, .NET, Python, Microsoft Graph API
Technologies: Vertex AI, Gemini Live API, voice agents, LLM guardrails, tool/function calling, prompt engineering, LangChain, LangGraph, RAG, vector databases, Model Context Protocol (MCP), Claude Code
Technologies: AWS (Lambda, SQS, SSM Parameter Store, EventBridge, EC2, Beanstalk, S3), GCP, Docker, Docker Compose, CI/CD (GitHub Actions, Azure DevOps, Jenkins), PostHog feature flags
- Microservices, event-driven architecture, Monolith, MVC, Design Patterns
- Trunk-based development, feature flags, expand-and-contract migrations
- Messaging: AWS SQS, RabbitMQ
- Testing: TDD, BDD, Unit & Integration Testing (Jasmine, Jest, JUnit)
- Building with the EU AI Act in mind
- Shipping trustworthy, regulation-aware AI features in production
- Agentic workflows with LangGraph and MCP
- Real-time voice agents: guardrails, tool calling and prompt design
- Learning projects:
- Email: mumarm45@gmail.com
- Location: Ingolstadt, Bavaria, Germany
- Phone: +49 162 473 9773
- LinkedIn: linkedin.com/in/mumarm45
- Website: www.omarmuneercoder.de
⭐️ Open to conversations on Full-Stack, AI Engineering & DevOps
Last Updated: October 2026