About Broker LoopBroker Loop builds agentic AI for insurance brokers that removes manual workflows and speeds up quoting, intake, and follow up. Our platform integrates real-time voice, SMS, email, and internal systems to capture compliant data, trigger next steps, and keep teams in flow.
The roleYou will design and build production systems across our voice AI and automation stack. The work spans real-time telephony and audio streaming, LLM-driven orchestration, prompt and experiment management, and the web dashboards that brokers use every day. You will ship pragmatically, measure outcomes, and improve reliability and latency with each release.
What you will doBuild and operate low-latency services for real-time calling, STT and TTS, and conversation control using TwilioImplement agent with Langchain, manage prompts and experiments with Langfuse, and track quality metricsCreate API endpoints and event-driven workers to handle quoting, data capture, and third-party integrationsDevelop front-end features in React and TypeScript for case review, call replay, audit trails, and reportingInstrument services for observability, logging, and cost tracking, then use the data to drive improvementsWrite maintainable code, review PRs, and improve our CI builds and deployment workflowCollaborate with founders, brokers, and partners to turn field feedback into shippable increments
What you will bring5 plus years of software engineering in production environmentsStrong skills in at least one of Go or TypeScript plus solid API design skillsExperience with cloud infrastructure, Docker, and CI on GitHub ActionsComfort with real-time or streaming systems such as WebSockets or telephony or audio pipelinesClear thinking about reliability, idempotency, and stateless service patternsPractical testing mindset plus a focus on observability and measurable outcomesExcellent communication and a bias to ship
Nice to haveTwilio ConversationRelay or Voice experienceDeepgram or ElevenLabs or similar audio stackGroq or other LLM providers and experience tuning latencyPostgres and Redis and event buses such as SQS or NATSSecurity and compliance awareness in financial services
Our stackBackend Go, Node and TypeScriptFront-end React and TypeScriptLLM and Langchain for agent, Langfuse for prompt and experiment managementTelephony and audio Twilio, Deepgram, ElevenLabsInfra AWS, Docker, GitHub Actions, TerraformData Postgres, RedisObservability OpenTelemetry, logs to Grafana
Ways of workingSmall units with clear ownership and thin slicesTrunk-based development with short-lived branches and PRsPragmatic tests and strong instrumentationWritten design notes for risky changesSecurity first mindset and customer trust as a core value
BenefitsMelbourne preferred with remoteNew hardware and tools that help you move fastLearning budget and access to advanced AI toolingCompetitive salary
Hiring processShort intro call with a founder to align on mission and rolePractical exercise that mirrors our real work, time boxed and reasonableTechnical deep dive and system design conversationTeam fit and offer