2 months ago
Responsibilities
- Define the architecture for the analyst AI stack across prompts, tools, metadata, memory and context strategies, evaluation, and runtime safety.
- Build abstractions and interfaces that let product teams add agent capabilities without brittle prompt sprawl.
- Design retrieval and context strategies balancing cost, latency, and quality.
- Establish evaluation loops, benchmark suites, and review processes to measure, debug, and improve model behavior.
- Partner with product, engineering, and analyst-domain experts to translate analytical workflows into AI-assisted product capabilities.
- Drive quality standards for correctness, trust, transparency, and failure handling in analyst-facing AI experiences.
Requirements
- Deep experience building AI-enabled product systems rather than only running experiments.
- Strong software architecture skills involving APIs, contracts, orchestration, and system boundaries.
- Practical understanding of LLM prompting, tool use, retrieval, structured outputs, and evaluation methods.
- Ability to work across backend engineering, product constraints, and user workflow design.
- Sound judgment about where agent systems should be flexible versus constrained.
- Comfort working in ambiguous environments where patterns are still being invented.
Benefits
- Full-time hybrid position based in Oakland, California, with two days in the office each week.
- 100% employer-paid medical insurance, subject to country and worker-type variations.
- Generous paid time off, paid sick time, inclusive parental leave, holidays, a year-end Global Week of Rest, and volunteer days off.
- RSU stock grants.
- Professional development and training opportunities.
- Virtual happy hours, free food, and team-building activities.
- Monthly cell phone stipend.
- Access to a mental health support platform offering therapy, coaching, and mindfulness resources for covered employees and dependents.
Categories
About Fivetran
Fivetran builds a managed data integration platform that extracts data from hundreds of SaaS apps and databases and loads it into cloud data warehouses for analytics and AI teams. Sold as a subscription SaaS service, it provides prebuilt connectors, automated schema management, and ELT/replication to destinations like Snowflake, BigQuery, Databricks, and Redshift. Founded in 2012 and headquartered in Oakland, it serves customers such as Autodesk, Condé Nast, JetBlue, and Morgan Stanley.