Automation & AI work companion

Your own mini tools

Somewhere in your business there's that one Excel file that holds everything together – and that everyone hopes nothing happens to. Or the process there's "surely some software for" that then has 80 functions too many.

Your own mini tools

01Your own mini tools

Mini tools live exactly in between: small, custom applications for one concrete process. An overview that shows project status. A capture tool for working hours on site. An analysis that sorts incoming inquiries. Since AI helps with the building, tools like these take shape in days instead of months – as a clickable prototype first, with real sample data. You try it out and then decide: use it in everyday work, extend it or scrap it. The basis for the decision costs days, not budgets.

Straight up: prototype and product are two different things. A tool with critical data, many users, rights and real-time demands is real software development with its own planning. The full build is therefore never offered directly, but only once the small version has proven itself in everyday use.

02Demo

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

03How it works

  1. 01

    Choose the one process

    Cut small and clear: one process, one pain, one tool. What lives in the heroic Excel file today is usually the perfect candidate.

  2. 02

    Clarify data and users

    Which data flows, who works with it, who may see what? For an internal three-person tool that's answered quickly – but it's always answered.

  3. 03

    Build the prototype in days

    A clickable version with sample data – close enough to the real thing that you can judge whether the tool hits your everyday work.

  4. 04

    Test with real cases

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

  5. 05

    Decide deliberately

    Use it, extend it or scrap it – all three outcomes are fine. A prototype scrapped after four days is cheaper than any wrong purchase.

04Stages

The tiers here are the sales logic itself: it only ever moves on once the previous tier has proven itself.

Tier 1

Prototype

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

Tier 2

Internal production tool

Real data, simple access, a small user group – the tool moves into everyday work.

Tier 3

Extended tool

Roles, integrations, export, a maintenance agreement – only once Tier 2 carries everyday work.

05Technical implementation

  • Frontend: a small web app or an internal dashboard – built quickly with Astro/React or similar, deliberately plain in design.
  • Data to suit the case: Airtable or Google Sheets for a fast start, SQLite/Postgres/Supabase when it gets more structured.
  • Access kept simple: password protection, a magic link or your existing workspace login.
  • AI as a function where it helps: summarising, classifying, text suggestions, search – not for its own sake.
  • Export to where you need it: CSV, PDF, email, CRM or project tool.
  • Hosting: Vercel/Netlify or internal – prototypes sometimes even run locally until the decision falls.

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