Forward Deployed Engineering
A forward deployed engineer is a software engineer who works with your team, inside your business, instead of building software from a distance. They learn how the work really gets done, remove the steps that waste time, and automate the rest in the tools you already use.
Technical reader, or ready to start? See our AI implementation services
What a forward deployed engineer does
You stay in charge of every important decision.
A forward deployed engineer is a software engineer who works inside your business, learns how the work really gets done from the people doing it, and builds the fix into the software you already use.
“Forward deployed” means they come to where the work happens instead of building from a distance. The role was made famous by Palantir, and AI companies such as OpenAI and Anthropic now have forward deployed engineers of their own. We bring the same approach to businesses of 10–100 people.
Almost every business knows this one. A customer fills in your website form or sends an email, and someone has to read it, research the company, check the CRM, write a proposal and remember to follow up. Here's that process before and after we redesign it.
Illustrative example: sales inquiry automation
Typical problems
Each step is tagged by who does it: simple rules, AI, or a person.
A person stays in control of the pricing and the final proposal.
| Before | After | |
|---|---|---|
| Hands-on time per inquiry | 20–30 minutes of manual work | 2–5 minutes of review |
| Tools opened by hand | Several: inbox, CRM, the company's website, past proposals | None: the details come to you |
| Response time | Sometimes hours | As soon as you approve the draft |
| CRM updates | Depend on who handled the inquiry | Automatic, every time |
| Proposal | Written from scratch or copied from the last one | A draft ready immediately, for you to check |
Most businesses try AI by buying a tool or some licenses and hoping it helps. People often like it, but the business runs the same way, because the tool sits on top of the old process. If the process is full of waiting, chasing and re-typing, AI just does the same steps a little faster.
Follow one supplier invoice through a typical business. The actual work adds up to 18 minutes, but the invoice takes 12 days to be ready to pay.
The rest of the time it's waiting: in an inbox, for a reply, for a signature. Speeding up the work barely helps. Removing the waiting is what makes the difference.
That's why a forward deployed engineer looks at the whole process, not just one task.
Illustrative example: one supplier invoice
18 minutesof actual work
12 daysfrom arriving to ready to pay
Sits down with the people who do the work and asks how it really happens, including the exceptions nobody wrote down.
Combines those conversations with the history in your software to show every step, every handoff and every delay. For most owners it's the first time they've seen the whole process on one page.
Decides what to remove, what simple rules can handle, where AI can help, and where a person must stay in charge.
The automation works inside your email, accounting software, CRM and Slack or Teams, so nobody has to learn a new system.
Measures the process before and after, so you see the difference in numbers, not opinions.
Two illustrative examples: sales inquiries and supplier invoices. Pick one and click through the four stages, from how the process looks on paper to how it runs after the redesign.
Illustrative example: sales inquiries at a 30-person services firm
How most businesses describe their sales process: five tidy steps, no exceptions.
What really happens, found by talking to the team and looking at the history in their CRM and inbox: more steps, back-and-forth, waiting, and leads that slip through.
Each step gets one of four answers: remove it, handle it with simple rules, let AI help, or keep a person in charge.
The new process. Everything is prepared for you: you check the pricing and the proposal, approve it, and the rest happens on its own.
This is the heart of the job. Each step of your process is either removed, handled by simple rules, given to AI, or kept with a person.
Steps that only exist because of how work gets passed around: renaming files, copying into spreadsheets, forwarding emails, chasing people. The cheapest automation is a step you don't need.
In the examples
Steps that follow a fixed rule, like “invoices over $5,000 also go to the finance director”. Ordinary software handles these: fast, predictable and cheap.
In the examples
Steps that need some judgment but follow patterns from the past, like reading an invoice in any format or picking the right account code. AI handles them, and anything it's unsure about goes to a person.
In the examples
Decisions that move money, commit the business or can't be undone. A person always makes these, with everything they need already in front of them.
In the examples
Automation shouldn't mean losing control. Anything that moves money, commits the business or can't be undone stays with a person.
The difference is that the work of preparing the decision is done for you, so a decision that used to take an afternoon of digging through emails takes a minute.
Invoice needs your approval#finance-approvals · just now
Harbor Packaging Co. · INV-20417 · $6,480.00
Why it's here
Billed for 40 cases, but only 36 were received. It's also over the $5,000 limit.
Evidence attached
Suggested decision
Approve $5,832.00 for the 36 cases received, and request a credit note for the rest.
Approve suggested amountRejectSend to someone else
We measure the process before we build anything, then measure it again after. Here's what that looks like for the invoice example.
Illustrative example: supplier invoices at a 40-person distributor
| Before | After | |
|---|---|---|
| Steps in the process | 13 | 6 |
| Time from invoice to ready to pay | 12 days | Same day, or 2 days for exceptions |
| Hands-on time per invoice | 18 minutes | About 2 minutes |
| Invoices that go straight through | 0%: every invoice is handled by hand | About 75%; people see only the exceptions |
| Times work gets sent back | 3 | 1 |
| Invoices fixed at month end | 12% | Under 2% |
Your numbers will be different. They come from your own process, measured before we build anything.
The same five phases for every client, whether you start with one process or a whole department. You approve each phase before the next one starts.
Phase 1 · Map
We interview the people who do the work, review the history in the systems involved and read the documents you already have. Then we measure where the time goes.
You get: The real process map, hands-on time compared with waiting time, the exceptions, and baseline numbers to measure against.
Phase 2 · Redesign
Every step is removed, handled by simple rules, given to AI or kept with a person. Together we agree the new process, who owns each step and what it should save.
You get: The redesigned process, the expected savings and a fixed quote. Nothing is built until you approve it.
Phase 3 · Build
We build the automation into the software your team already uses, test it on real data and run it alongside your team until it's reliable.
You get: Working automation, a short walkthrough for your team and simple documentation. Typically 2–6 weeks per process.
Phase 4 · Prove
After go-live we report the same numbers we measured in the first phase, so you can see exactly what changed.
You get: A before-and-after report you can share with your team.
Phase 5 · Improve
With Automation Care we monitor the automation, fix issues, update the AI as models change and start on the next process when you're ready.
You get: Automation that keeps working as your business changes.
The engineer who maps your process is the one who builds it, so nothing gets lost between a consultant and a developer.
We work in your CRM, accounting software, inbox and Slack or Teams. No new platform, and no migration first.
Every phase is quoted up front. If something isn't worth automating, we'll tell you.
We sign an NDA before we see any data (plus a BAA or DPA where you need one) and never use your data to train AI models.
What it costs
A forward deployed engineer is a software engineer who works inside a client's business instead of building software from a distance. They learn how the work really gets done from the people doing it, remove the steps that waste time, and build the automation into the software the business already uses.
Forward deployed engineering is a way of delivering software and AI where the engineer works alongside the client's team instead of taking requirements from a ticket. The same engineer maps the process, redesigns it, builds the automation into the existing systems and measures the result, so nothing gets lost between a consultant and a developer.
“Forward deployed” means the engineer works where the work happens, with the client's team and inside their systems, rather than back at the vendor's office. The term was made famous by Palantir, which sent engineers to work on site with its customers, and AI companies such as OpenAI and Anthropic now have forward deployed engineers too.
At Brainium it has five phases. Map: find out how the process really runs. Redesign: remove steps, and decide what rules, AI and people each handle. Build: put the automation inside your existing software and run it alongside your team until it's reliable. Prove: measure the result against the baseline taken at the start. Improve: keep it working and move on to the next process.
A developer or agency usually builds what you ask for. A forward deployed engineer starts by finding out what's really happening, often things nobody wrote down, and fixes the process before building anything. The result is simpler work, not just faster work.
No. You tell us where the time goes and introduce us to the people who do the work. We handle the technical side and explain everything in plain language.
Usually not. We build inside the tools you already use, such as your email, accounting software, CRM and Slack or Teams. Nobody has to learn a new system.
That's rarely the goal. Automation takes away the copying, chasing and re-typing, so your people can spend their time on customers, decisions and growing the business. Many businesses use the time to take on more work without hiring more admin staff.
We plan for it. AI only handles steps where its work is easy to check, rules double-check the results, anything it's unsure about goes to a person, and every action is logged. Decisions that move money or can't be undone always stay with a person.
A first look at one process usually takes days and is free for qualifying businesses. Building the automation for one process typically takes 2–6 weeks, at a fixed price agreed before we start. We never bill by the hour.
Yes. We sign an NDA before we see any data (plus a BAA or DPA where you need one), access is limited to what each automation needs, and we never use your data to train AI models.
Tell us what your team spends too much time doing. We'll map it, show you what can be removed or automated, and tell you whether the numbers make sense.
Find What You Can AutomateFree assessment. No commitment. If something isn't worth automating, we'll say so.
