SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Aleph is an AI-native Financial Planning & Analysis (FP&A) platform backed by top-tier VCs including Khosla Ventures, Bain Capital Ventures, and Y Combinator. The company was founded by technical founders from Stanford and MIT with experience at Google, Microsoft, and Bain & Company, and serves customers like Webflow, Notion, Zapier, and Y Combinator.
As Staff Software Engineer for AI, you will own the foundational AI systems that all product teams build upon. Your responsibilities include:
- Owning the AI foundation layer: model selection and routing, the model proxy layer, context management, and tool design
- Building and maintaining eval infrastructure that encodes what excellent FP&A output looks like, enabling all teams to measure quality (critical since incorrect answers can end up in board decks)
- Shipping agentic features end-to-end, from v0 through optimization: prompt engineering, context engineering, caching, parallel tool calls, and subagent patterns
- Creating observability into agent behavior to identify bottlenecks and drive data-informed prioritization
- Establishing best practices for agentic systems and bringing proven patterns into how the company builds
You should have shipped production LLM systems in evals, context management, multi-model routing, or similar domains. You build quickly, validate with evals, and ship v0s. You have strong judgment about abstractions—opinionated on design but pragmatic about incremental shipping. You're experienced with LLM APIs and agent frameworks and have shipped user-facing products.
As a Staff+ engineer at Aleph, you'll demonstrate adoption (teams you don't manage use systems you built), strong fundamentals across the stack, high agency with a builder's mindset, pragmatism in tool selection, craft in code quality, clear communication across audiences, and experience building complex B2B products ideally from scratch. This is a remote-first role open across the Americas (US, Canada, LATAM) to align with team working hours.
