AI Backend Engineer
Posted 46ds ago
Employment Information
Report this job
Job expired or something wrong with this job?
Job Description
AI/Backend Engineer developing AI orchestration pipeline for financial technology. Owning the AI Pipeline and shaping technical decisions for a scalable platform.
Responsibilities:
- Own the AI Pipeline
- Design and optimize our multi-agent orchestration system
- Implement parallelization and streaming to dramatically reduce response latency
- Build robust prompt management with versioning and A/B testing capabilities
- Build RAG Systems
- Design retrieval-augmented generation for accurate, contextual responses
- Work with vector databases, embeddings, and relevance scoring
- Optimize for both speed and accuracy at scale
- Develop Production APIs
- Build developer-friendly APIs connecting our AI capabilities to the frontend
- Design for future integrations with CRMs and advisor tools
- Implement proper authentication, rate limiting, and documentation
- Shape the Foundation
- Establish code review practices and testing standards
- Document architecture decisions for future team members
- Contribute to technical patents and IP development
Requirements:
- 4+ years production Python experience (async patterns, type hints)
- Hands-on experience with LLM APIs (OpenAI, Anthropic, or similar)
- Strong understanding of prompt engineering and multi-step LLM workflows
- Production API development experience (FastAPI or similar)
- Strong SQL and PostgreSQL skills
- Great to Have: Experience with RAG systems and vector databases (Pinecone, Weaviate, pgvector)
- Streaming/real-time implementation experience (SSE, WebSockets)
- TypeScript/JavaScript familiarity
- FinTech or regulated industry background
Benefits:
- Equity: Meaningful early-stage grant with 4-year vesting
- Equipment: Professional Laptop ready to work with AI provided + stipend for remote work when 6 month mark is met
- Time Off: Flexible PTO with a minimum 15 days encouraged
- Learning: $1,000 annual professional development budget
- Schedule: Flexible hours with 3-4 hours daily overlap (Americas timezones)













