San Francisco, California, United States of America
Full Time
3 weeks ago
$150,000 - $220,000 USD
No Visa Sponsorship
Key skills
PythonTypeScriptAIMLAnalyticsDatadogCaching
About this role
Role Overview
You'll work on live audio systems: buffering, streaming, latency optimization, and the integration points with speech providers.
You'll build and improve the systems that make conversations feel natural: dynamic knowledge injection, action execution with branching logic, and the dialogue management that turns a phone call into something useful.
You'll work on the action framework: configurable API calls, success/failure branching, auth management and the runtime execution engine.
You'll work on how agents ingest, store, and retrieve knowledge at runtime.
You'll work with our designers to translate cutting edge advanced AI concepts and features into smooth, reliable and easy to use product experiences.
You'll help build evaluation frameworks for each model involved, post-call analytics, call quality metrics, and Datadog instrumentation.
Requirements
You have experience building production backend systems ideally in Python or Typescript (all levels welcome, 5+ years for senior roles).
You write async code naturally, understand how to design for performance and reliability, and have shipped systems that real users depend on.
You understand what makes real-time systems different from request/response API’s. You can talk about the tradeoffs around buffering, streaming, connection management, and latency.
You've worked with LLMs or AI/ML systems in production. You understand concepts like prompt engineering, context management, function calling, and what it takes to make AI behavior reliable for customers.
You know evaluation matters and you've built or used tooling around it.
Event driven architectures, message queues, caching layers, and async task processing feel natural to you.
You've debugged distributed failures and designed systems that handle partial failures gracefully.
Testing strategy, observability, API design, and code review are things you do thoughtfully. You write code that other engineers can understand and extend.
You communicate clearly. You can explain a complex technical problem to someone without your context, write a design doc that prevents misalignment, and ask good questions before diving into implementation.
You are familiar with AI coding tools and already embedded them into your day to day engineering processes.
Tech Stack
Python
TypeScript
Benefits
Medical, dental, and vision insurance is 100% covered
401k plan with company matching!
Unlimited PTO — take the time you need to come to work feeling great!