Aircall.io, Inc. · posted · checked
Software Engineer, AI Voice Agent
PythonTypeScriptOn-site · MadridSenior
Who can apply
On-site · Madrid
Location: Madrid Office
Aircall is a unicorn AI-powered communications platform transforming how teams win with customers.
For over a decade, we’ve believed voice builds the strongest business relationships. We bring voice, SMS, WhatsApp, and 250+ integrations together on one seamless workspace—so every customer conversation drives revenue, resolves issues faster, and builds trust. Our AI Voice Agents handle routine interactions while your team leads the conversations that matter most. Used by 24,000+ businesses worldwide.
We’re 600+ people from 40+ nationalities across Paris, London, Madrid, Berlin, New York, San Francisco, Seattle, Sydney, and Mexico City. We’re ambitious, product-driven, and execution-focused—built for visible impact, fast decisions, and real growth.
Come join us!
About the role
Businesses running on Aircall field calls their teams can’t handle alone, such as after hours, at peak volume or across languages. AI Voice Agent answers them. It is not an IVR or a canned response bot. It is a real-time voice agent that understands context, talks naturally, executes actions mid-conversation and knows when to hand off to a human. The role sits at the intersection of real-time systems, production LLM applications and voice infrastructure, where milliseconds matter and quality is “immediately audible.” This is “production AI at scale,” not prototyping.
What you’ll work on
Real-time speech pipeline: Live audio buffering, streaming and latency optimization, and evaluating or integrating speech providers.
Conversation intelligence: Prompt construction, context management, function calling, dynamic knowledge injection and dialogue management.
Actions and integrations: The action framework, covering configurable API calls, success/failure branching, auth management and the runtime execution engine.
Knowledge and memory: How agents ingest, store and retrieve knowledge at runtime, and retain information across a conversation.
Agent lifecycle and configuration: Working with designers so customers can create, configure, test and deploy agents easily.
Evaluations and quality: Evaluation frameworks per model, post-call analytics, call quality metrics, Datadog instrumentation and the on-call rotation.