Forward Deployed Engineer – Applied AI & Search

About Chaps Vision & Sinequa

We build the enterprise platform for AI-powered search and agentic orchestration. The goal is straightforward: give every employee fast, governed access to the right information, in the right context, when they need it, with tight control over how AI uses it.

As part of the North America Professional Services team, you'll be the Forward Deployed Engineer (FDE) embedded on-site with one of our largest and most strategic customers, a global manufacturing leader running a multi‑year co‑innovation program to put agentic AI in the hands of its workforce 's a hybrid role: business lead and technical lead in one. You'll own the translation of the customer's priorities into Sinequa delivery, and of platform capability into outcomes its teams actually adopt.

You'll build production‑grade applications, help shape the platform roadmap from the field, and work across the customer's stakeholders, its embedded delivery team, and our R&D group.

A lot of forward‑deployed roles give you a set playbook and a rotating list of accounts. This one is set up differently.

You'll own a single flagship account from start to finish, building against real production data (drawings, specifications, parts, PLM) rather than demo environments. And because this is the deployment that shapes how we scale, what you work out here will become the blueprint for the next wave of strategic customers. The strongest parts of what you build can make their way into the core platform roadmap, so your field work reaches every customer we have.

You'll have real ownership of the account, with the platform, the R&D relationships, and company resources behind you.

• Embed with the customer: Sit on‑site at the customer's headquarters, working day‑to‑day with its business stakeholders, its own program team, and technical leadership. You'll be the main point where account strategy and technical delivery come together.
• Translate in both directions: Turn business priorities into a delivery plan you can actually build, and turn platform capability into outcomes that work for the people using them. A lot of the value here is judgment and translation, not just code.
• Lead technical discovery: Run the architecture discussions, whiteboarding, and discovery needed to map a complex data ecosystem across PLM, drawings, specifications, and parts data spanning several business units.
• Resolve blockers at the desk: Work through a rigid enterprise environment with high interpersonal instincts. You're comfortable walking over to a stakeholder's desk to unpack an integration roadblock and get people to agreement.

• Ship production artifacts: Write clean, production‑grade code for complex components like Model Context Protocol (MCP) servers, stateful sub‑agents, and deterministic agent skills on our agent orchestration platform.
• Model‑agnostic design: Build workflows that use the right frontier LLM for the job (Claude, Gemini, GPT, or specialized open‑source models), tuned for cost, context limits, and latency.
• Optimize search at scale: Manage high‑scale indexing and search pipelines for a user population in the tens of thousands. That means search‑engine performance work, document chunking, and real‑time relevancy tuning across a range of data silos, including the PLM‑sourced technical content (for example, PTC Windchill) that sits at the core of this program.
• Build in strong guardrails: Design clear process guardrails and required Human‑in‑the‑Loop (HITL) checkpoints, including SME validation, role‑based access, and a capability maturity lifecycle, so autonomous agents working in low‑margin‑for‑error settings stay safe, predictable, and auditable.

• Bridge PS and R&D: Sit between Professional Services delivery and core R&D. Carry escalation authority and roadmap visibility, and route what you find in the field into structured product feedback.
• Accelerate the handover: Bring the product depth that helps the customer's embedded team take over more of the work over time. Part of doing this job well is gradually…