Skip to content
Sortup

How we work

Small first steps, working software every two weeks.

Progress is a staging link, not a slide deck. Everything is yours from the first commit.

A build, week by week

Example
An example six-week build: scope in days 1–2, a written update every week, a staging link every two weeks, code in your repository from the first commit, and handover in the final week. Example of a six-week build. Product builds run 6–12 weeks.

How a project runs.

  1. 01 · 20 minutes

    Fit call

    You bring
    The problem, in your words, and who it is for
  2. 02 · Days 1–2

    Scope

    You bring
    Access to the people, data and systems involved
  3. 03 · 2-week milestones

    Build

    You bring
    Feedback on each staging link, and a decision-maker who replies
  4. 04 · Final week

    Launch and handover

    You bring
    The people who will run it

Three ways to work together.

Start with a pilot, then choose a build or a pod. Stop after any milestone.

5–14 days

A pilot or audit

A Knowledge Assistant, an Agent Pilot or a Production Audit.

  • Written scope in days 1–2
  • Live, documented and handed over
6–12 weeks

A product build

The web or mobile product around the AI, in 2-week milestones.

  • Production release, monitoring and handover
Month to month

A dedicated pod

2–4 engineers plus a lead, for roadmaps that do not end at launch.

  • The same people every sprint
  • Your repository, your backlog, your tools

8 commitments, all of them checkable.

Engagement commitments08 items
Engineer reply
Within 1 business day
Code and repository
Yours from commit 1
Staging link
Every 2 weeks
Eval set
Agreed on day 1
Written update
Every week
NDA
Yours or ours, before the call
After launch
We fix what we shipped, 30 days
Handover
Runbooks, docs and a call
What each commitment means
  1. 01An engineer replies within one business dayNot a sales rep. The person who would scope the work.
  2. 02You own the code from the first commitYour repository, your cloud account, your API keys.
  3. 03A staging link every two weeksYou click through working software, not slides.
  4. 04An eval set agreed on day 1For AI work, we agree how "good" is measured before we build.
  5. 05A written update every weekWhat shipped, what is blocked, what is next.
  6. 06We sign your NDAOr send ours, before the first call if you want.
  7. 07We fix what we shipped for 30 days after launchBugs in delivered work are ours to fix.
  8. 08Every project ends with a handoverRunbooks, architecture notes and a call, so you can run it without us.

Your data

What happens to your data.

Diagram: your data, index and API keys sit inside your cloud account. The model provider is reached through API calls under terms that exclude training. Sortup engineers get least-privilege access, removed at the end of the engagement.
  1. 1We use the model providers you choose: OpenAI, Anthropic, Google or open-weight models you host.
  2. 2We use provider APIs under terms that exclude training on your inputs and outputs.
  3. 3Your data and indexes live in your cloud accounts wherever possible; we get least-privilege access for the project.
  4. 4Access is removed when the engagement ends, and we sign your NDA and DPA.

For overseas buyers

We work your hours.

The engineers on your project keep your working day: stand-ups, calls and replies happen while you are at work.

Your working day, and ours

Standard time
  • GCCDubai · UTC+410:00–19:00in Lahore
  • UKLondon · UTC+014:00–23:00in Lahore
  • US EastNew York · UTC−519:00–04:00in Lahore
The engineers on your project work your 09:00 to 18:00. In Lahore that is 10:00 to 19:00 for the GCC, 14:00 to 23:00 for the UK, 19:00 to 04:00 for the US East.

Paperwork

NDA
Yours or ours, before the call
IP and code
Yours from commit 1
DPA
We sign yours

Every commitment

Questions people actually ask.

You are in Lahore. Will you work our hours?

We work your hours: the engineers on your project keep your working day, whether you are in the GCC, the UK and EU or the US. Stand-ups, calls and replies happen in your day, not ours.

Who will we talk to?

The engineer leading your build, on every call. No account manager in between.

How do we know the code is any good?

A second engineer reviews every change and CI checks it. You can read the repository yourself from day one.

Who owns the IP?

You do, from the first commit. It lives in your repository, and the contract says so.

What happens after launch?

30 days of fixes on what we shipped. Then you run it with our runbooks, or keep a small pod month to month.

AI makes things up. How do you handle that?

We measure it: an eval set agreed on day 1, citations on answers drawn from documents, and human approval before anything irreversible.

What happens to our data?

It stays in your accounts wherever possible, under model-provider API terms that exclude training on it, and access is limited to the people on your project. We sign your DPA.

We tried an offshore agency before and it went badly.

Then you know the failure modes: silence, surprises and code nobody can read. Our answer is a written update every week, a staging link every two weeks and your repository from day one.

Want to see if we fit?

No pitch deck. An engineer, and a straight answer.