3 days ago
Berlin, GermanySenior
Responsibilities
- Build reliable voice AI systems with RAG, dynamic LLM routing, and shadow testing for new transcribers.
- Design and operate a distributed call scheduler supporting concurrent calls, load balancing, race-condition handling, retries, time zones, and provider limits.
- Extend agents to calls, WhatsApp, and email with shared memory and state across asynchronous and realtime conversations.
- Build memory and reasoning systems that preserve context across customer flows.
- Develop an AI coworker that configures agents, debugs issues, and integrates with CRMs, calendars, and email.
- Improve tracing, root-cause analysis, and resilience when calls, workers, or providers fail.
Requirements
- Production experience building and owning distributed systems at scale, including race conditions, load balancing, and fault tolerance.
- Strong proficiency in Python with FastAPI and/or TypeScript with Node.js, Bun, or Fastify, plus end-to-end software delivery experience.
- Hands-on production experience with LLMs, AI models, or agentic systems.
- Familiarity with PostgreSQL, Redis, or similar production data stores.
- Experience with AWS, GCP, or Azure infrastructure.
- Familiarity with Temporal, message queues, or similar workflow orchestration tools.
- Experience with OpenTelemetry, Honeycomb, or equivalent observability tooling.
- Realtime systems, audio processing, or voice infrastructure experience is strongly preferred.
- React or full-stack experience is a welcome addition.
- Ability to work independently, challenge assumptions, drive outcomes, and collaborate with a team focused on business impact.
Benefits
- €85,000 to €160,000 annual salary, depending on experience.
- Equity included.
- Visa sponsorship available.
- Relocation support is available for candidates moving to Berlin.
- Fully on-site in Berlin, Germany; remote work is not available.