About Autonex
At Autonex, we exist so people can get back to working like people. We free customer service and operations teams at large companies from repetitive communications through results-oriented conversational AI capable of managing long-running processes. Our voice agents handle thousands of real calls in sectors such as healthcare, sports, and logistics, so people can focus on what truly adds value. We are looking for an intern who wants to learn how to build voice agents for enterprise environments: at scale, with complex integrations, and with the reliability required in production.
Responsibilities
•Build voice agents in Python with LiveKit Agents and LangGraph.
•Integrate them with APIs and client systems so they can perform real tasks.
•Compare STT and TTS providers for latency, accuracy, and naturalness.
•Analyze calls with LangSmith to detect failures and propose measurable improvements.
Requirements
•Be able to sign an internship agreement.
•A good foundation in Python, APIs, and Git.
•Know how LLMs work: prompting, tool calling, and structured outputs.
•Have experimented with an agent or voice framework, even in personal projects.
•Curiosity about voice AI and a mindset of measuring before considering something good.
Nice to have
•A repository or personal demo, experience with Claude Code, or an interest in agent evaluation (WER, LLM-as-a-judge, A/B tests).
Conditions
•Learn alongside the founders and a Full-Stack professional with more than 15 years of experience.
•Real impact from the first month.
•Flexible, hybrid 6-hour workday.
•Office in Sant Andreu, Barcelona.
•Your birthday off.
