SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 145,000 - 165,000 / annual
Prenuvo is transforming healthcare from reactive sick-care to proactive health through award-winning whole-body MRI scans (fast, safe, non-invasive) combined with cutting-edge AI. We are seeking a Senior Forward Deployed AI Engineer to embed with clinical, product, and operations teams to ship production-grade AI features that are validated, governed, and monitored by default.
In this role, you will own AI solutions end-to-end—from intake brief through shipped, monitored production features. You will make critical build/wrap/vet decisions that set validation bars to keep unvetted features away from patients. You'll build agentic layers including tool and function calling, RAG, MCP integrations, multi-step orchestration, and human-in-the-loop checkpoints. You will architect solutions on the Prenuvo AI platform (shared model serving, foundation-model contracts via AWS Bedrock, eval harness, monitoring, tool registry, prompt guardrails) with no parallel stacks.
Key responsibilities include defining evaluation criteria before building (acceptance criteria, offline/online evals, escalation triggers), running initiatives through AI intake and risk categorization so governance is inherited rather than bolted on, and leaving reusable assets behind each engagement (validated templates, registered tools, reference architectures). You will raise AI fluency across teams through design reviews and office hours, enabling other teams to ship AI features independently from a template rather than a blank page.
You'll work closely with senior clinical, product, and engineering stakeholders, translating fuzzy product requests into validated designs and scoping initiatives with sound product delivery instincts.
**Requirements:**
- Six or more years building and delivering product-grade software with a track record of shipping AI-powered features to production (LLM and/or imaging using tool calling, RAG, or agent orchestration). Notebooks and demos are not sufficient; you must be an experienced full-stack developer.
- Probabilistic-software literacy: ability to evaluate models, reason explicitly about failure modes, and answer performance questions with methodology rather than guesswork.
- Judgment and standing to hold a validation bar with senior clinical, product, and engineering stakeholders, explaining reasoning in their language.
- Product delivery instincts: scope fuzzy requests, say no to wrong versions, translate product needs into validated designs.
- Strong Python and practical cloud engineering/deployment (APIs, deployment, ideally AWS and Bedrock); comfort working directly inside another team's codebase.
- Hands-on with modern evaluation frameworks and human-in-the-loop design; you've built eval harnesses and acceptance criteria, not just consumed them.
**Nice to Have:**
- Experience in regulated, clinical, or healthcare-data environments (FDA, Health Canada, ISO 13485, or equivalent) and comfort with PHI.
- AWS and Bedrock specifically (though any major cloud translates).
- MCP or experience building tool/function interfaces other teams consume.
- Consulting, solutions architecture, or forward-deployed experience (embedded in another team before).
