Natera is a company focused on enhancing efficiency and productivity using AI and automation. They aim to leverage advanced technologies, including AI agents and integrations, to accelerate workflows across various business domains.

About the Position

Introduction

The Forward Deployed Solutions Engineer at Natera will embed within specific business domains to identify, build, deploy, and operate AI agents and automation pipelines. This role focuses on enhancing efficiency and productivity by leveraging AI to accelerate workflows, bringing significant impact to various functions like Commercial, Clinical Operations, and Sales & Marketing.

Responsibilities

Find the leverage in your domain:
- Map existing and potential workflows to identify opportunities for AI, ML, or automation to improve throughput, coverage, or speed.
- Develop business cases, quantify projected impacts, and align priorities with domain leadership.
Design the future-state workflow:
- Map data flows (structured and unstructured) across various systems (CRM, ERP, ticketing, document stores, internal tools, external SaaS).
- Define target workflows, clearly delineating agent responsibilities, human roles, and hand-off points.
- Determine necessary context for agents or models (retrieval, grounding, tool access, memory) and ensure reliable delivery.
- Design human-in-the-loop checkpoints to add value without creating bottlenecks.
Build and connect the systems:
- Implement agents and automation pipelines using approved AI platforms and frameworks.
- Integrate agents with business systems via MCP servers, APIs, webhooks, CLIs, and skills, adhering to IT and security guidelines.
- Configure tools, prompts, context, and retrieval pipelines for reliable performance in real-world scenarios.
- Address integration challenges such as authentication, schema drift, rate limits, and data quality.
- Provide access and training to business users for operating these workflows.
Run agents and automation pipelines in production:
- Maintain end-to-end ownership of agent performance, tracking KPIs like throughput, quality, cost, human intervention rate, and adoption.
- Develop and manage evaluations, re-running them for any significant model, data, or workflow changes before deployment.
- Troubleshoot failures, refine prompts and context, iterate on workflows, and decommission agents when no longer suitable.
- Implement observability tools, including tracing, structured logs, and dashboards, to ensure transparency.