Forward Deployed AI Solutions Engineer

About the role The Forward Deployed Solutions Engineer will work directly within a business domain (e.g., Commercial, Clinical Operations, Lab Operations, Sales & Marketing, Customer Experience etc.). In your role, you’ll find opportunities for enhancing efficiency and productivity by looking for workflows which can be executed 10–100x faster or more often than a human team could using AI agents, integrations and other patterns. You will build, deploy, and run them in production. You report into the central AI & Automation team, partner directly with domain leadership on priorities, and bring patterns back so the whole company compounds. Find the leverage in your domain Map the workflows in your domain — the ones running today, and the ones that don’t exist yet because they weren’t feasible without agents or automation tools. Identify the step-change opportunities: where AI, ML, or automation unlock throughput, coverage, or speed. Build the business case, quantify projected impact, and align with domain leadership on priorities. Design the future-state workflow Map structured and unstructured data flows across the systems involved (CRM, ERP, ticketing, document stores, internal tools, external SaaS). Define the target workflow: what the agent does, what the human does, and where they hand off. Figure out what context the agent or model needs to do the work well — and how to get it there reliably (retrieval, grounding, tool access, memory). Design human-in-the-loop checkpoints so review adds value without becoming the bottleneck. Build and connect the systems Stand up agents and automation pipelines using the organization’s approved AI platforms and frameworks. Connect agents to business systems — via MCP servers, APIs, webhooks, CLIs, and skills — within the guardrails set by IT and security. Configure tools, prompts, context, and retrieval pipelines so agents perform reliably on real work, not just in demos. Handle integration gnarliness: auth, schema drift, rate limits, data quality, and the messy last-mile of enterprise systems. Enable access and training for business to run the workflows Run agents and automation pipelines in production Own agent performance end-to-end. Track the KPIs that matter — throughput, quality, cost, human intervention rate, cycle time, adoption. Build and manage evals. Re-run them on any material model, data, or workflow change before it ships. Triage failures, tune prompts and context, iterate on the workflow, and retire agents when they’re no longer the right tool. Instrument observability: tracing, structured logs, dashboards. You don’t ship what you can’t see. What we’re looking for Hands-on technical fluency. CLIs, APIs, webhooks, SQL, and Python scripting. Working knowledge of LLM and agent behavior — prompting, context, tool use, RAG, MCP, evals, failure modes. Be very comfortable with a cloud platform. Trustworthy with elevated access. Least-privilege, auditability, and safe rollbacks are second nature. Strong technical and process judgment. You think in outcomes and KPIs, can defend prioritization calls, and are comfortable being the most technical person in a business meeting and the most business-savvy in a technical one. Nice to have Prior experience working hand in hand with businesses to deliver measurable outcomes. Hands-on experience with an enterprise agentic platform (CrewAI, LangChain, AWS Bedrock, Claude, Codex) or building directly against a model API. Background in product management, solutions engineering, consulting, forward-deployed engineering, or technical operations. Experience in regulated environments (HIPAA, SOC 2, GxP, SOX). Success in year one Shipped three or more workflows into production that are measurably moving a business KPI, with agent evals and observability in place. Domain leadership brings you into planning early, not late. Contributed at least one reusable asset another engineer on the team is now using. Enabling non builders to become builders using t