Forward Deployed Engineer

About the role

This role is based in Prague (3 days/week in the office). You own a client's problem from the first conversation to the thing running in production, and then you find the next problem. As a Forward Deployed Engineer at Keboola, you sit with our clients, figure out what will actually move their business, design the solution, build it with your team, and make sure it gets used.

Keboola Professional Services is a small team inside Keboola that builds solutions for new clients and helps existing ones get more out of the platform. Forward Deployed Engineers own the client relationship and the solution design. Data Engineers own delivery. Both build.

We build agent-first. Coding agents write most of the code. You own the system around them: the architecture, the context agents need to get a client's business right, and the checks that prove the result is correct. More and more of what we deliver is itself for agents: governed data, definitions and skills that let a client's people hand real work to AI.

This role is for someone who's equally at home in a CFO's office and in a terminal, and who would rather ship a working solution than write a recommendation about one.

What you'll do: Win the case

• Lead the conversation at the top: sit with CFOs, COOs and CEOs, understand what's on their agenda, and shape what data and AI should deliver for it
• Work out with their teams how the business runs and where the data lives, then design the solution and scope it into a statement of work the client can sign and we can deliver
• Run proofs of concept together with Data Engineers, on the client's own data, against success criteria agreed up front
• Own the engagement
• Be the client's main technical counterpart, from kickoff to go-live, with business stakeholders and data teams alike
• Make the architecture and design calls, and build hands-on alongside Data Engineers on the parts where your judgement matters most
• Engineer the context: capture the client's metric definitions, business rules and ownership so that people, reports and AI agents all get the same right answer
• Keep the project on scope and budget, and raise trade-offs and scope changes early
• Grow the account
• Stay close to the executives sponsoring the clients you've launched: review how the solution is used, spot the next use case, and turn it into the next project
• Help clients put agents to work on top of their data, so their own teams can delegate real work to AI
• Work with Sales and Customer Success on account plans for the clients you own
• Make the team smarter
• Turn what you learn on one client into reusable patterns, agent skills and context for the next one
• Feed concrete client needs and platform gaps back to Product and Engineering

What we're looking for: Must-haves

• Executive presence.
• You've advised C-level executives directly, not through a partner or a manager. You can walk into a CFO's or COO's office, understand what's on their agenda, connect it to what data and AI can actually do for them, and leave with the next meeting booked.
• Client ownership.
• You've been the person a client relied on to get a data or software project from problem to production. You can turn a vague ask into a scoped plan and say no to the wrong thing without losing the room.
• Solution design.
• You can design data solutions end to end: sources, modelling, transformations, consumption by reports, apps and agents. You know the trade-offs and can explain them to a technical and a business audience.
• Agentic engineering.
• You build with coding agents every day. You work in tools like Claude Code, connect agents to platforms through MCP, and break work down so agents can do it. You read what they produce well enough to catch a wrong join or a silently dropped row, and you know when to take over.
• Context engineering.
• You know an agent is only as good as its context. You can capture a messy business, what its numbers mean, who owns them and which rules apply, in a form both people and agents can work from.
• English.
• Professional working proficiency. You'll work with clients across Europe and the US.
• Strong pluses
• Finance and accounting data: general ledger, chart of accounts, consolidation, management reporting
• ERP, CRM or accounting system integration (NetSuite, SAP, Dynamics, Salesforce and similar)
• Pre-sales or solutions experience in a data or SaaS company
• Building AI agents and LLM use cases for clients, not only using agents to build
• Cloud data warehouses: Snowflake, BigQuery or Databricks
• Nice to have
• Consulting, system integrator or startup background
• Czech
• How we measure success: We tell you up front what good looks like.
• New clients live.
• Clients you're staffed on go live on the date agreed in the statement of work, within budget, and sign off on what they got.
• Accounts grow.
• The clients you own use more of the platform a year after you've worked with them, and come back with the next project. Their executives treat you as an advisor and bring you into their next decision.
• Pre-sales.
• The opportunities you lead turn into signed clients, and the solutions you scope are the ones that get delivered.
• Time with clients.
• Most of your time goes to client work: discovery, design, delivery and account growth.
• Team.
• The patterns, skills and context you contribute make the next engagement better.
• Who thrives here: Experience: 5+ years in data, analytics or software, with at least a few of them in front of clients, including direct work with C-level executives.
• Background: Many paths lead here. You might be a data engineer who became the person clients asked for, a solutions engineer who wanted to build what you demoed, a consultant who got tired of handing over slides, or a finance person who learned to build. Domain depth counts as much as code: if you understand how a business runs, we'll take that seriously.
• Mindset: You'd rather own the outcome than the hours. You're comfortable with ambiguity, you make calls and record them, and you care whether the client is using what you built a year later. You treat agents as part of the team: you delegate to them, check their work, and keep improving how they work.