Process

How a project with us runs

Six steps, no mystery. The same shape whether it's an MVP, a mobile app or a voice agent.

  1. 01

    Scoping call

    20–30 minutes

    A short, direct conversation about what you're trying to achieve. We ask questions, you ask questions, and we tell you honestly whether we're the right studio for it.

    What you get

    • An honest read on scope, timeline and fit
    • A recommendation — even if it's not us

    What we need from you

    • Your idea, in your own words
    • Any existing product, deck or notes you have
  2. 02

    Written cut list and proposal

    2–4 days

    We cut the product down to the smallest version that proves real value, then put it in writing: what's in, what's out, what it costs, when it ships.

    What you get

    • A written cut list — what's in and what's out
    • A fixed price and a timeline, signed off by you

    What we need from you

    • One decision-maker who can say yes or no
    • Feedback on the cut list within a few days
  3. 03

    Design

    1–2 weeks

    Screens and flows for the core journeys, designed before code so we argue about pixels, not about rebuilt features.

    What you get

    • Clickable designs of the core flows
    • A design pass with your feedback baked in

    What we need from you

    • Your brand assets, if you have them
    • Fast, honest feedback on the designs
  4. 04

    Build in weekly milestones

    Weeks, not months

    You see working software every week — a link you can click, not a status deck. Progress updates land every 48 hours, and feedback goes straight to the engineers building your product.

    What you get

    • A working build to click every week
    • Progress updates every 48 hours
    • A say in trade-offs as they come up

    What we need from you

    • Feedback within a day or two on milestones
    • Access to any accounts or systems we integrate with
  5. 05

    Testing and launch

    Ship week

    We test the real journeys end to end, deploy, monitor, and stay close through the first real users. Post-launch fixes are part of the deal, not an extra invoice.

    What you get

    • A tested, deployed production release
    • Two weeks of post-launch fixes included

    What we need from you

    • Your final go / no-go
    • Real users ready to try it, if you have them
  6. 06

    Handover and support

    Your call

    You own the code, the docs and the accounts. Hand it to your own team, or keep us on a monthly retainer to keep shipping. Either way, no lock-in.

    What you get

    • Code, cloud accounts and domains in your name
    • Handover docs your next developer can use
    • Optional monthly retainer to keep shipping

    What we need from you

    • A decision: hand over, or keep building together

Voice AI projects

The voice-specific variant

Voice agents run on the same principles, with a different shape: script before code, and a pilot on real calls before launch.

  1. 01

    Define the call goal

    One conversation

    We pick exactly one call to automate — renewals, bookings, lead qualification — and define what a successful call looks like.

    What you get

    • A single, sharp call goal
    • Success criteria in writing

    What we need from you

    • Recordings or notes of real calls, if you have them
    • The person who handles these calls today
  2. 02

    Script and knowledge base

    2–4 days

    We write the call script and assemble the approved knowledge the agent may use — answers, prices, policies — and you sign off on every word.

    What you get

    • A call script you approve line by line
    • A knowledge base the agent can't go outside

    What we need from you

    • Your real answers to the questions callers ask
    • A decision-maker's sign-off on the script
  3. 03

    Build and test with sample calls

    1–2 weeks

    The agent is built and wired to your calendar, CRM or backend — then we hammer it with sample calls, in every language it will speak.

    What you get

    • A working agent on a test number
    • Sample-call recordings to review

    What we need from you

    • Access to the systems it integrates with
    • Ears — your feedback on how it sounds
  4. 04

    Pilot with real calls

    2–3 weeks

    The agent takes real calls in a controlled slice — after-hours only, one cohort, one city — while we review every transcript together.

    What you get

    • Real-call transcripts and outcomes
    • Language quality tested on real phone audio

    What we need from you

    • A pilot scope you're comfortable with
    • Fast feedback when something sounds off
  5. 05

    Tune and launch

    Ship week

    Pilot learnings go in, the guardrails are confirmed, and the agent goes live on its full scope — with a human fallback always one step away.

    What you get

    • A launched agent on your own number
    • Handover docs and a review dashboard

    What we need from you

    • Your final go / no-go
    • A human fallback contact for handoffs
  6. 06

    Monthly review

    Ongoing

    Every month we review outcomes and cost per call together, tune what drifts, and decide what the agent takes on next.

    What you get

    • A monthly outcome and cost report
    • Continuous tuning from real calls

    What we need from you

    • Thirty minutes a month
    • Honesty about what's not working

Step one is a conversation

Twenty minutes, no pitch deck. Bring your idea and we'll tell you what we'd build first — and what we'd skip.

Book a call See our work
Book a call