AI Discovery Sprint

Know where to start with AI. In two weeks.

Work with our senior engineers to discover where AI could improve your business and connect your teams, systems and data. Together, we choose where to start.

Martín PrunellEngineer & Co-Founder

Talk with Martín about your business and whether the sprint is a good next step.

Book a 30-minute intro call

What we work out together

Where could AI help?
Opportunities to improve processes and connect work across teams and systems.
What would it take?
The data, integrations and effort involved, with the risks still to check.
What should happen first?
A first scope to build, or the preparation needed before you start.

Two-week sprint · Fixed price

How it works

Two weeks, working with your team.

Your team shows us how the business works. Together, we choose the areas to explore, then assess the opportunities with the engineers who would build them.

What we need from you: a person who can approve the next step, 3–4 hours per week from the people involved, and read access to the relevant systems or their documentation.

Days 1–2

Show us the work

We look at how work moves between your teams and systems, using real tasks to understand the handoffs and recurring problems.

Days 3–5

Explore the opportunities

We examine the available systems and data, then map opportunities to improve processes and connect work across the areas we agreed to explore.

Days 6–8

Check the assumptions

Compare the options with us: likely impact, effort and what still needs evidence.

Days 9–10

Choose a first step

Review the scope, estimates and open questions together. If something is missing, the roadmap says what to prepare.

Our work

Engineering experience behind the sprint

You can review nine years of core product engineering at Auth0 and Frankie, the voice assistant built for Amgen’s field sales reps.

Sophilabs presented us with a product that we were proud to show people.
Noel GomezData Science Lead, Amgen

Four deliverables for the first build.

What you walk away with

  1. 01

    AI opportunity map

    The workflows we reviewed and the opportunities worth investigating.

  2. 02

    Prioritized use cases with ROI estimates

    Impact and cost estimates, with assumptions and the evidence still needed.

  3. 03

    Technical feasibility & data requirements

    Data, access, integrations, and unresolved technical questions.

  4. 04

    Pilot definition & execution roadmap

    First scope, acceptance criteria, milestones, and budget assumptions.

Illustrative example Imagine a woodworking workshop

Orders arrive through WhatsApp and email. Designs live in 3D tools. Materials and machines have their own records. What could work better if these systems shared information?

  • Orders & suppliers

    WhatsApp, email and purchasing tools.

  • 3D design

    Models, dimensions and revisions.

  • Materials

    Stock, costs and availability.

  • Workshop machines

    CNC and cutting software.

Find the connections

An approved design could feed a materials list. Stock and supplier information could help plan the job. Updates from the workshop could inform the team handling the customer’s order.

Check the existing tools

  • What can each tool share through an API, an export or a file?
  • Can we match orders, materials and design revisions across systems?
  • Which decisions need a person’s approval, and who has access?

Choose a first connection

For one type of job, bring the order, approved design and materials availability into one workspace. AI could flag missing information and prepare a plan for review.

People approve materials and production instructions. Direct machine control stays outside this first scope.

Discovery would confirm access, compatibility and effort before we propose what to build.

When Discovery is useful

Before you start

A good fit

  • You have a business or product in operation.

    Your team can show us how work happens today, using real tasks, software and data.

  • You want to explore where AI could help.

    We look across teams and systems to find useful connections and choose where to start.

  • Someone can decide what to build next.

    A person from your team can weigh the options and agree on a first scope.

Not the right fit

  • You already know exactly what to build.

    Talk to an engineer about starting implementation directly.

  • There's no software or data to work with yet.

    The sprint needs an existing process and data to assess. It is too early to book one if your project is still an idea.

  • You need a general report on AI.

    The sprint assesses opportunities in the areas we agree to explore and defines a first implementation scope. It does not cover general market research.

Fixed scope · Fixed price

2 weeks

Fixed price, credited 100% toward implementation contracted within 90 days.

We agree on the scope and price before starting. The four deliverables are yours to use with your team or another supplier.

If the work is not ready to build, the roadmap explains the gaps and the next decisions.

After the sprint

What happens next is your call

Your team can use the roadmap to carry out the work. If you continue with us, we can quote a fixed price for the first use case or a monthly subscription based on agreed deliverables. The scope and budget estimates help you decide.

Build with a forward-deployed engineer

Frequently Asked Questions

How much does an AI discovery sprint cost?
We quote a fixed price after a scoping call. The agreed sprint fee is credited 100% toward implementation if you contract that work with us within 90 days.How every engagement is priced
Can you add AI to an existing SaaS product without rebuilding it?
We assess how the feature can fit your permissions, data model, interface, and release process. The feasibility work identifies the integrations and any changes your existing software would need.
Who runs the sprint?
The senior engineers who would build the work. You review the findings and the proposed scope directly with them.
What happens if we decide not to build?
You keep everything: the map, the prioritized use cases, the data assessment and the pilot definition. It is scoped so another team can execute it.
How does an engagement start?
Tell us how your business works and what you would like to improve. We help identify opportunities and agree what to explore during the sprint. If the scope is already defined, we can start with implementation.
Why is it paid when others offer free AI assessments?
You pay for two weeks of engineering work and four deliverables you own. You can take them to your team or another supplier. If you contract implementation with us within 90 days, the sprint fee is credited in full.
Is it remote?
Yes. The sprint runs remotely, with sessions scheduled in your time zone.
What about confidentiality and our data?
We sign a non-disclosure agreement (NDA) before the first technical conversation. We work with the access you grant and follow your security controls. The Trust Center explains our access practices, data handling, and ownership of the work.Trust Center
Can our own team execute the roadmap?
Yes. The roadmap includes the scope, assumptions, and acceptance criteria for your team to use. If you later contract implementation with us within 90 days, the sprint fee is credited in full.
What if our data isn't ready for AI?
The feasibility assessment identifies data gaps and the preparation needed. That work may become the first step of the roadmap.

Tell us what your team wants to improve.

Explore what could work better when your teams, systems and data connect. We can help you find where AI would be useful and choose a first step.