SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Abacus Insights is transforming how data works for health plans by making healthcare data usable so decision-makers can act faster with confidence. The company helps health plans break down data silos to create a single, trusted data foundation that powers better decisions, improved outcomes, reduced waste, and better member and provider experiences. Backed by $100M in funding, the platform enables GenAI use cases through clean, connected, and reliable healthcare data.
The Senior Forward-Deployed AI Engineer is a senior individual contributor responsible for designing and building AI and GenAI/agentic systems that power the platform on Snowflake, with a significant portion of time spent working directly inside client environments. This is not a purely internal engineering role; candidates must bring demonstrated prior experience doing forward-deployed or client-facing implementation work.
Initially, the role focuses on configuring the Snowflake semantic layer and delivering metrics and reporting for clients. As that foundation matures, the work expands into agentic AI systems that depend directly on the semantic layer. On client engagements, this person serves as the senior-most AI technical voice, trusted to work through ambiguous problems and help shape technical vision and roadmap for AI components. This is technical leadership without people management or engagement ownership responsibilities.
Day-to-day responsibilities include: designing and configuring Snowflake semantic layers (e.g., Cortex Analyst semantic views) with accurate business logic and metrics definitions; building and delivering metrics and reporting off the configured semantic layer; working directly with client stakeholders to define requirements, validate outputs, and resolve discrepancies; writing production-grade SQL and Python for semantic layer configuration and metrics pipelines; designing and implementing RAG systems, agentic workflows, and MCP-based tooling; building evaluation harnesses to test and monitor AI/agent quality; serving as the senior technical voice on client engagements to scope AI/ML use cases and translate business requirements into technical solutions; implementing and configuring solutions within client environments; troubleshooting issues live with client teams; and partnering with product, data science, and operations teams to move AI/ML work from prototype to production.
The role blends AI/ML architecture and model strategy on Snowflake, GenAI and agentic systems design, and direct hands-on client engagement and implementation. Unlike purely internal engineering roles, this position regularly puts you in front of clients, scoping needs, implementing solutions in their environment, and troubleshooting alongside them.
REQUIREMENTS:
- 8+ years of software/AI engineering experience, including hands-on design of AI/ML systems
- Deep, hands-on experience with Snowflake, including Snowpark, Snowflake Cortex, and Snowflake ML
- Hands-on experience building and maintaining semantic layers on Snowflake (e.g., Cortex Analyst semantic views), including modeling business logic and metrics for AI/agent consumption
- Experience delivering metrics and reporting off a semantic layer for clients, including validating outputs against business logic
- Demonstrated experience designing and building GenAI/agentic systems: RAG, MCP-based tooling, agent orchestration patterns, and evaluation harnesses
- Proven track record working directly with external clients or customers in an implementation, forward-deployed, or professional-services capacity, with specific engagements personally owned (not only internal engineering experience)
- Demonstrated ability to operate as the senior technical voice on ambiguous problems, with sound judgment on AI technical direction, without needing close direction
- Strong Python engineering skills, including API design and delivery of production AI/ML pipelines
- Strong communication skills with demonstrated ability to translate technical tradeoffs for non-technical, client-facing audiences
- Comfort operating with ambiguity typical of client environments, and sound judgment on when to escalate versus resolve independently
NICE-TO-HAVE:
- Hands-on experience with Databricks, including Unity Catalog, Delta Lake, and MLflow
- Healthcare payer or provider claims data experience
- Experience with classical ML and model fine-tuning
- Relevant certifications (e.g., SnowPro)
