Tailored Tools

Tiny software for real problems.

I started Tailored Tools to fix the tedious, repeated tasks no off-the-shelf product handles. I build a working tool for the people who have the problem, and I ship it fast. Every app here is live and in regular use.

Start with the annoyance

Each tool begins with a chore someone does by hand, on a schedule, and prefers not to.

Small on purpose

Built for a handful of users and scoped so it never has to hold sensitive personal data.

Measured in time saved

Success means the chore takes minutes, or seconds, instead of an evening.

Selected tools

Each one is live and in use today
01

Bid: Found

Every week, 600+ municipal bid sites checked automatically and the matching bids in one list
Runs every Monday

The problem

My dad consults for utility companies across the US. To win work, he has to find open bids that match what he does, and every municipality posts its bids on its own website, in its own format.

He had no structured way to keep up. He spent a few hours each week clicking through sites looking for anything relevant, and still reached only a fraction of the municipalities he cared about. Any bid he missed was work he never got the chance to pursue.

The old way

  1. Remember which municipal sites to check
  2. Visit each one and find its bids page
  3. Skim every listing for relevant work
  4. Run out of time well before the list ends

Who uses it

One consultant, my dad, every week.

How Bid: Found works

A scheduled job scrapes 600+ municipal bid sites each week, which takes about 15 minutes, and matches every open bid against his keywords. When he logs in on Monday morning, the matching bids are waiting for him to review.

In a typical week, it scans more than 2,000 open bids across 12 states and narrows them to a few dozen worth his attention.

Weekly, automatic
Scrape 600+ sites

Runs on a schedule. About 15 minutes, with no one watching.

Automatic
Match keywords

Keeps only the open bids that fit his work.

Monday morning
Review matches

He decides which bids to pursue.

Coverage every week

600+ municipalities checked every week, with each dot one site

Where it could go

Any consultant or contractor who chases public-sector bids has this same problem across hundreds of sites.

Built with

About 5 days from start to live, including discovery across the municipal sites. Built with Claude Code.

  • Python 3.11
  • FastAPI
  • SQLAlchemy
  • Jinja2
  • PostgreSQL
  • httpx + BeautifulSoup
  • Playwright
  • APScheduler weekly job
  • Docker on Railway
  • Daily GitHub Actions backup

His time spent finding bids, per week

Before: checking sites by handa few hours
Hoursbefore, partial coverage
After: reviewing Monday's matches~5 min
5 minafter, 600+ sites
600+ sites checked weekly 100+ hrs of searching saved per year 2,275 → 43 bids scanned vs. matched in one weekly run
What he sees on Monday: matching bids grouped by state, tagged with the keyword that caught them, with due dates.
Every run is logged, with per-state results and errors, so a broken source is easy to spot.
His 23 keywords can be added, paused or removed in the app, with no code changes.
02

Clurb

Monthly book selection, from nominations to winner, in three clicks
In use monthly

The problem

I organize a monthly book club. Every month, picking the next book meant chasing nominations, pasting them into an email, sending it out for a vote, and tallying the replies by hand. It was about an hour of admin every month, and it all ran through my inbox.

The old way

  1. Collect nominations from members
  2. Compile them into one email
  3. Send it out and wait for votes
  4. Tally replies and announce the pick

Who uses it

The 7 members of my book club, every month.

How Clurb works

Members submit their own nominations and cast their own votes. As the organizer, my whole job is to move the club from one phase to the next.

Members can nominate straight from their Goodreads want-to-read list or search for any book. Nobody can vote for their own pick, and the results show the vote count and who leads the discussion.

Phase 1
Open nominations

Members add the books they want to read.

Phase 2
Open voting

Members vote on the list. No tallying by hand.

Phase 3
Close voting

The winner is announced to the club.

Built with

About 8 hours from idea to hosted V1. My first app, built with Claude Code.

  • React 19
  • Vite
  • Node.js
  • Express 5
  • SQLite (better-sqlite3)
  • Railway

Organizer time per month

Before: email and manual tally~60 min
60 minbefore
After: three clicks in Clurb~30 sec
30 secafter
99% less organizer time ~12 hrs → ~6 min per year 0 votes counted by hand
The full monthly cycle in about 30 seconds, with pauses cut: open nominations, members nominate from Goodreads or search, open voting, members vote, announce the winner.

Approach

Founder

Alexandra Richardson

I build each tool myself, with Claude Code, for the people who will use it.

Most of the time lost at work and at home goes to small, recurring chores that no commercial product fits well enough to fix. I build for those gaps. A tool used by a handful of people every month can still give back hours a year to each of them.

I use Claude Code to go from a problem to a working, hosted app in hours or days rather than months. That speed lets me test whether an idea is useful with the people who have the problem, then refine it based on how they actually use it.

Keeping each tool narrow is a deliberate decision. Clear scope keeps the apps easy to maintain and keeps sensitive data out of them, and the same judgment carries over to larger products.

Have a tedious process worth fixing?