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 the artifacts you created.