Cloud DB Architect / ML & Agentic AI Engineer Knuff & Kunde CPAs Eagle, ID Job Details Part-time | Full-time | Contract 6 hours ago Benefits Health savings account Health insurance 401(k) Flexible schedule Qualifications AI models Data model design Data compliance Continuous Delivery (CD) implementation Data retrieval time improvement Software implementation AI platforms (beyond public GPTs) Machine learning projects Schema design Infrastructure architecture design Production systems Cloud database architecture Docker Access control implementation Data access controls implementation APIs Developing and maintaining backend systems Integration with Third-Party Services (system development) Cloud compliance Linux Reducing data breach risk by implementing security protocols Data Loss Prevention DevOps automation AI-driven automation Query management Linux administration Python Data security architecture Financial compliance Full Job Description We're building proprietary AI software for our accounting firm, transitioning our internal tools into a multi-tenant SaaS platform. Independently funded with no outside investors — you'll have the runway to build this the right way, not the fastest way. You'll join our developer and UI/UX designer to own the cloud architecture, data infrastructure, and agentic AI systems end-to-end. What You'll Own Agentic AI Workflows Design and harden autonomous AI agents (Vertex AI, Claude) that reliably execute multi-step financial workflows — with validation layers, fallback logic, and guardrails against hallucination and context drift. Cloud Infrastructure Architect the migration from a local environment to a secure, scalable, multi-tenant cloud deployment ready for external firms to onboard. Data Layer Design and optimize PostgreSQL schemas handling complex financial data alongside vector-based AI memory (pgvector) — with zero tolerance for data integrity issues. Required Skills AI Systems Production experience building agentic workflows against Vertex AI and/or Claude APIs Strong grasp of RAG architecture and retrieval strategies Practical techniques for constraining LLM outputs — structured outputs, tool-use validation, state tracking across multi-turn tasks, and recovery/fallback logic when a step fails or hallucinates Backend & Data Python 3.10+, FastAPI, AsyncIO at a production level PostgreSQL schema design for relational + JSONB data, query optimization, pgvector implementation and tuning Cloud & DevOps GCP experience strongly preferred Docker, Linux administration, CI/CD pipeline design Comfortable owning infrastructure decisions with minimal oversight Security "Secure by design" development practices Encryption at rest/in transit, RBAC implementation Experience architecting data-loss prevention boundaries — specifically around what does and doesn't leave your infrastructure when calling external AI APIs Familiarity with compliance considerations for sensitive PII (financial data experience a plus) To Apply Send your resume, GitHub/portfolio, along with brief answers to: How do you engineer agentic workflows against external APIs (Vertex/Claude) so they reliably execute multi-step tasks without hallucinating or losing context? How do you balance rapid MVP deployment against the strict data privacy requirements of sending financial PII to an external AI provider?

Benefits:

401(k) Flexible schedule Health insurance Health savings account

Work Location:

Hybrid remote in Eagle, ID 83616