01
Discovery (week 0–1)
Free 15-minute call. If there's a fit, paid discovery sprint to scope the agent against your real data and workflows.
// SVC.FDE
Forward-deployed AI engineering — production agents shipped into your stack.
Foundation models are commodified. The real bottleneck is enterprise SSO, legacy ETL, regulated data, and your security review board. We bridge it. Two to three engineers embed with your team for 4–12 weeks, build the integrations, deploy the agent, and stay on retainer until it pays for itself.
We don't sell SaaS. You own the code, the prompts, the integrations. We hand you the keys when it's stable and remain on call for tuning. This is the Palantir FDE model — minus the seven-figure minimum.
timeline: 4–12 weeks build · ongoing retainer optional pricing: $30K–$100K build · $500–$5K/month retainer
WHAT'S_INCLUDED
The deliverables
01
Discovery + architecture: what problem the agent solves, where it integrates, the kill criteria
02
Production-grade integration with your existing stack (Google Ads, GA4, GSC, GitHub, your CRM, etc.)
03
A custom-tuned agent with hard limits, audit trail, and approval gates configured for your team
04
Control surfaces: Slack bot, your existing chat, or our custom UI — pick one or all
05
Documentation, runbooks, on-call coverage during ramp-up
06
Source code + prompts + infrastructure-as-code transferred to your repos
PROCESS
How it runs
02
Build (week 2–8)
Two-engineer team embedded with your stakeholders. Weekly demos. Production deploy by week 8 unless the integrations are genuinely huge.
03
Hand-off + retainer (week 8+)
Code transfer, runbook, training. Optional ongoing retainer for tuning, new capabilities, and on-call.
OTHER_PATHS
Not the right fit? Maybe one of these.
READY.TO_TALK
Let's see if this is the right shape.
15-minute discovery call. We'll tell you if we're the right fit or point you to someone who is. No pitch deck.