Forward deployed AI engineer
Not a slide deck. Not a hand-off. I work inside your business, find where time and money are leaking, and ship working AI into your operations — then I stay until your people run it.
A forward deployed engineer works inside a customer’s business, finds the problem that matters most, and builds and ships the software that fixes it. Siamak Kalhor works this way for small and mid-size businesses: an AI consultant with 40 years of marketing and business strategy who also builds and deploys the AI agents and software himself.
What a forward deployed engineer does
The role was made famous by Palantir, whose engineers sat with customers instead of shipping software from a distance. AI companies such as OpenAI and Anthropic now hire forward deployed engineers for the same reason: a powerful model is worth little until someone puts it to work inside a real business, with real data and real people.
The job has three parts. Embed — learn how the work is actually done, not how the org chart says it is. Find the constraint — the one place where leads, hours or money are lost. Ship — build the fix, put it into production, and measure whether it worked.
Why a strategist who builds
Most AI projects fail before the first line of code: they automate the wrong thing. Forty years in marketing and business — from growing Blue Moon Fabrics roughly 100× to advising law firms, medical practices and manufacturers — is what decides what to build. The engineering decides whether it actually works on a Monday morning.
You deal with one person from the first conversation to launch. There is no account manager in between, and the person who diagnoses the problem is the person who ships the fix.
How an engagement runs
- Embed (week 1). I sit with your calls, your inbox, your tools and the people doing the work. Remote or on site in Los Angeles and Napa Valley.
- Find the constraint. Working back from your unit economics, I name the bottleneck worth fixing first — and tell you plainly if the answer is not AI.
- Ship. A working agent, app or automation in production in weeks, not quarters. You approve every customer-facing answer before launch.
- Measure and hand over. Your team is trained, the system is monitored and tuned, and you own the accounts, data and content.
What gets shipped
Recent systems built and deployed this way:
- Lending Desk — an AI front desk for private lenders that takes borrower calls, qualifies deals and routes them to the lender.
- Your Legal Brain — AI intake for law firms: answers, qualifies and summarises every inquiry before a lawyer picks up.
- Ledger Desk — an AI voice front desk for accounting firms in four languages that scores every lead and puts the recorded call, transcript and summary in the firm’s inbox.
- Site Manager — a voice-driven field app for a Napa Valley custom-home builder to log hours, jobs and crews hands-free.
- AisleBird — a voice-first grocery app that sorts lists by store aisle, live across 300+ locations.
More in the portfolio and case studies.
Forward deployed engineer, consultant or agency?
| Dimension | Traditional consultant | Forward deployed engineer |
|---|---|---|
| What you get | A diagnosis and a recommendation | A diagnosis and working software in production |
| Where the work happens | In workshops and documents | Inside your operations, with your data and people |
| Who builds it | Someone else, later | The same person who diagnosed it |
| How success is measured | Deliverables handed over | Hours saved, leads captured, revenue moved |
For the full comparison with large agencies, see AI consultant vs full-service AI agency.
What it costs
Every price is published on the pricing page:
- $350 — one-hour strategy session, booked online
- $350–$450 / hour — consulting and engineering time
- $3,500 + $1,000 / month — done-for-you AI agent, including hosting, monitoring and tuning
- Fixed quote — larger custom AI software, priced before work starts
When this is not the right fit
- You need a standing development team or 24/7 contractual support.
- You are building an enterprise platform or your own AI model from scratch.
- You want a report to file away rather than a system your people will use.
Find the one thing worth fixing first
One hour with Siamak: your bottleneck, whether AI is the answer, and what it would take to ship.
Book a $350 strategy hour →Frequently asked questions
A forward deployed engineer works inside a customer's business rather than from a distance. They learn how the work is actually done, find the problem that matters most, and build and ship software that solves it in production. Palantir made the role well known, and AI companies such as OpenAI and Anthropic now hire forward deployed engineers to put AI into customers' operations.
Yes, that is how he works with clients. Siamak Kalhor is an AI consultant who also works as a forward deployed AI engineer for small and mid-size businesses: he sets the strategy, builds the AI agents and software himself with his own AI tooling, deploys them into the client's operations, and stays until the client runs them. He is independent, not a forward deployed engineer employed by an AI lab.
A traditional consultant diagnoses the problem and recommends what to do. A forward deployed engineer also builds and ships the fix, inside the business, and measures whether it worked. Siamak Kalhor does both: 40 years of marketing and business strategy decide what to build, and hands-on engineering gets it into production.
Typically AI voice and chat agents that answer and qualify customers, intake and lead-management systems, industry apps, workflow automation, and AI-ready websites. Examples Siamak Kalhor has shipped include Lending Desk for private lenders, Your Legal Brain for law firms, Ledger Desk for accounting firms, and a field app for a Napa custom-home builder.
Prices are published. A one-hour strategy session is $350. Hourly consulting and engineering is $350 to $450. A done-for-you AI agent is $3,500 to set up plus $1,000 a month for hosting, monitoring and tuning. Larger custom builds are quoted as a fixed price before work starts.
No. That is the point of the model for small businesses. Siamak Kalhor does the building and deployment, trains the people who will use the system, and approves every customer-facing answer with you before launch.