Skip to content
PAKO Technology

Better systems. Bigger possibilities.

We work with teams whose operations have outgrown the tools that got them here. The data already exists. Nobody can act on it. And the answer isn't another dashboard.

tiroom

Every story that matters, read once.

tiroom reads Nepali and international coverage of the same story, removes the repetition, and gives you one brief you can actually finish.

  • Nepali and international sources, collapsed into one story
  • A brief you can finish before your coffee goes cold
  • A subscription opens the archive, topic alerts and full-text search

darta

Fill any form once. Never type it twice.

darta is a browser extension that completes forms from data you already hold — including forms it has never seen before.

  • Works on forms it has never encountered, with no per-site setup
  • Your records stay on your machine until you press submit
  • Correct a value once and every later form inherits the fix

How the work runs

Four stages, in order.

We don't skip to the build. Every stage exists because the one before it produced something the next one needs.

    Develop

    Work out what is actually wrong. Usually it is not the thing that got escalated. It is two steps upstream of that, in a process nobody has written down.

    Build

    Ship the smallest system that changes the outcome. Narrow enough to finish, real enough to run on live data rather than a sample.

    Scale

    Take it from working once to working every day, under load, with the edge cases your operation actually produces.

    Grow

    Hand it over properly. Your team runs it, extends it and fixes it without calling us. That is the only definition of finished we accept.

How we engage

We stay long enough to see it work.

Plenty of firms deliver on the last day of the contract and leave you the hard part, which is the first month in real conditions. We treat that month as part of the job. If the system does not survive contact with your operation, it was not built.

Writing

What we've been working out.

Notes on building systems that hold up, written while building them rather than afterwards.

  • 2 min read

    Another dashboard is not the answer

    Dashboards fail for a reason that has nothing to do with the dashboard. Usually the numbers disagree, and no amount of visualisation fixes a disagreement.

    • Systems design
    • Data
  • 3 min read

    What ranking a news feed actually costs you

    Every ranked feed makes an editorial decision and calls it an algorithm. Building tiroom meant deciding what we were willing to optimise for.

    • Product
    • Applied ML
  • 2 min read

    Why we don't skip to the build

    Every client who asks to skip the first stage has a good reason. They are still wrong, and the cost shows up in the same place every time.

    • How we work

Common questions

Tell us what's slowing you down.

Not a budget, not a company size. Just the thing that keeps coming back. We reply within two working days.