Automation & AI work companion

Your own mini tools

Somewhere in your business there's that one Excel file holding everything together – and everyone hopes nothing ever happens to it. Or the process there's "bound to be some software for", which then turns out to have 80 features too many.

You'll get …

  1. 1

    A tool for your process

    Instead of the Excel file holding everything together, or software with 80 features too many: a small tool that does only what you need.

  2. 2

    A clickable prototype in days

    With sample data, and close enough to the real thing for you to judge whether it fits the way you work.

  3. 3

    A clear decision, not a bad purchase

    Use it, extend it or scrap it – tested on real cases before anyone commits a budget.

01Intro

Mini tools fill the gap in between: small, custom-built applications for one specific process. Because AI helps with the build, tools like these come together in days rather than months – first as a clickable prototype with realistic sample data. You try it out and then decide: use it day to day, extend it or scrap it.

A prototype and a product are two different things. A tool with critical data, lots of users, permissions and real-time requirements is proper software development and needs its own planning. That's why the full build is never offered straight away – only once the small version has proved itself in day-to-day use.

02How it works

  1. 01

    Pick one process

    Keep it small and clear: one process, one pain point, one tool. Whatever lives in that all-important Excel file today is usually the ideal candidate.

  2. 02

    Clarify data and users

    What data is involved, who works with it, who's allowed to see what? For an internal tool used by three people, that's quickly answered – but it always gets answered.

  3. 03

    Build the prototype in days

    A clickable version with sample data – close enough to the real thing for you to judge whether it fits the way you work.

  4. 04

    Test it on real cases

    The last twenty inquiries, the current project list, a real week of time tracking: real material shows what's missing and what can go.

  5. 05

    Make a clear decision

    Use it, extend it or scrap it – all three outcomes are fine. A prototype scrapped after four days costs less than any bad purchase.

03Stages

Here the tiers are the sales logic itself: we only move on once the previous tier has proved itself.

Tier 1

Prototype

Clickable, with sample data, built in days – the basis for your decision.

Tier 2

Internal tool for daily use

Real data, simple access, a small group of users – the tool becomes part of everyday work.

Tier 3

Extended tool

Roles, integrations, exports, a maintenance agreement – only once Tier 2 is working well day to day.

04Try it yourself

simulation · curated examples

Three mini-tools from three completely different directions: one calculates for your customers (try it!), one shows the boss the week, one runs on the phone in the store. They have only one thing in common – each one is one process, one tool.

// profile

05Technical implementation

  • Frontend: a small web app or an internal dashboard – built quickly with Astro/React or similar, with a deliberately plain design.
  • Data storage to suit the case: Airtable or Google Sheets for a quick start, SQLite/Postgres/Supabase once things need more structure.
  • Simple access: password protection, a magic link or your existing workspace login.
  • AI as a feature where it helps: summarising, classifying, suggesting text, search – not for its own sake.
  • Export to wherever you need it: CSV, PDF, email, CRM or project management tool.
  • Hosting: Vercel/Netlify or in-house – prototypes can even run locally until a decision is made.

for context · This shows how something like this usually gets built – not how it has to be built for you. What actually fits is worked out in the project: around your existing setup, your team, and what stays easy to maintain.

06Related building blocks

See all use cases