Engagement model
Fixed-fee Sprint
One deliverable, one fixed price, one fixed timeline. The default.
When you know what you want shipped, a Fixed-fee Sprint is the cleanest path. We commit to scope and price after a 1–2 week discovery; you commit to the dependencies (access, sign-off owners, the data we need). Overruns are our problem, not yours.
Numbers
- Typical price
- $40K – $150K
- Typical timeline
- 2 – 8 weeks
- Discovery (paid)
- $8K – $20K · 1–2 weeks
- Payment schedule
- 30 / 30 / 40 · signing / midpoint / delivery
Best for
- A defined deliverable: 'build us a customer-support agent'
- Budget owners who need a fixed number before approval
- Teams who want a clean handoff with all IP transferred
- Most first-time engagements with us
Not for
- Exploratory work where you don't know what you want yet
- Multi-workstream builds across 3+ use cases (use Retainer)
- Engagements under $20K — overhead eats it
What you get
- Signed SOW with milestones and acceptance criteria
- Senior engineers in your repo from week one
- Working code at every weekly demo · not slides
- Eval suite + observability built in
- Runbook + on-call playbook at handoff
- All code, prompts, configs, IP transferred to your accounts
Payment terms
30% on signing · 30% at midpoint demo · 40% on delivery. Net-15 invoicing. Payable in USD, EUR, GBP, or INR. First-time clients usually 30% upfront on the first engagement, then standard.
FAQ
What if it takes longer than estimated?
Our problem, not yours. The point of fixed-fee is that you know the price before we start. Scope changes (you add something) are quoted as a separate add-on you approve before any work happens. Discovery quality is how we keep overruns rare — we don't quote until we're confident.
Can I add scope mid-sprint?
Yes — quoted as a discrete add-on, signed before work starts. We don't surprise-invoice.
What if the project doesn't ship?
Discovery is paid separately ($8–$20K, 1–2 weeks). If discovery surfaces a reason not to build, you walked away with that conclusion at 1/10th the cost. We've recommended 'don't build this' on 1 of every 8 engagements we've discovered.
Talk it through.
20 min. We'll tell you whether this is the right model for your situation — and if not, which is.