Job descriptionFull Time
Migaru AI builds an AI-native CRM that reads a sales team's email, calendar, and calls, then automatically files conversations, updates records, and drafts follow-ups and outreach for review. We're looking for an AI Engineer to help build and scale the agentic systems that keep the record current and do the work that usually falls through the cracks.
What you'll do
- Design, build, and improve the AI agents and pipelines that read conversational data (email, calendar, call transcripts) and extract structured, reliable updates to the CRM record
- Develop and refine LLM-based workflows for tasks like summarization, classification, deduplication, and judgement calls that require human review
- Build and maintain the retrieval and search infrastructure that lets agents answer questions accurately from a company's own records
- Evaluate and improve model quality, including building evaluation harnesses, test sets, and monitoring for hallucination, drift, and accuracy regressions
- Collaborate with product and engineering to ship agent-driven features such as meeting prep, sequence drafting, and account scoring
- Optimize for latency, cost, and reliability as usage scales across customer accounts
What we're looking for
- Professional experience building production systems that use LLMs, embeddings, or other ML models, ideally in an applied/product setting rather than pure research
- Strong software engineering fundamentals, including experience designing APIs, data pipelines, and services that handle real-world, messy input
- Hands-on experience with prompt engineering, retrieval-augmented generation, fine-tuning, or agent frameworks
- Comfort working with ambiguous, judgement-based problems, and building systems that know when to defer to a human
- Solid understanding of evaluation methodology for generative AI systems: how to measure quality, catch regressions, and iterate quickly
- Clear communication skills and comfort working cross-functionally with product and customer-facing teams
Nice to have
- Experience building systems that integrate with email, calendar, or calling platforms
- Experience with vector databases or other retrieval infrastructure at scale
- Background working on CRM, sales, or other go-to-market tooling
