← Back to the problem

The builder's path

Five steps from a problem to its solution.

For: People filing in NL courts can't quickly check that every case they cite actually exists.

“The true creator is necessity, which is the mother of our invention.”
Plato · Republic II, 369c (tr. B. Jowett)

Routes to a solution

Six ways to take this forward

Each route is a different kind of journey, with real people and organisations to contact at each step, starting with those in Newfoundland and Labrador.

  1. Step 1

    Build in the open

    Publish the code and invite contributors from the community.

    • Genesis ↗

      Newfoundland and Labrador's startup incubator: programmes, mentors and community.

    • techNL ↗

      NL's tech industry association: events, programmes and connections.

    • Code for Canada ↗

      Civic tech: technologists working with governments on public services.

    • Canadian Digital Service ↗

      Builds federal public services in the open; a route for problems with government services.

  2. Step 2

    Fund the maintainers

    Sponsors, collectives and open source grants.

    • ACOA ↗

      Federal funding and advice for Atlantic Canadian businesses.

    • Code for Canada ↗

      Civic tech: technologists working with governments on public services.

    • Canadian Digital Service ↗

      Builds federal public services in the open; a route for problems with government services.

    • NRC IRAP ↗

      Advice and funding for technology innovation in Canada.

  3. Step 3

    Get it adopted

    Civic tech groups and public bodies take it up.

    • ACOA ↗

      Federal funding and advice for Atlantic Canadian businesses.

    • NL Public Procurement Agency ↗

      How the Newfoundland and Labrador government buys; open calls are posted here.

    • Genesis ↗

      Newfoundland and Labrador's startup incubator: programmes, mentors and community.

    • techNL ↗

      NL's tech industry association: events, programmes and connections.

What this problem needs

Ideas, sources and tools for this problem, from first interviews to funding.

  • Web
  • Open data

Ideas to explore

  • Idea 1Citation pattern matcher plus lookupPull citations out with regular expressions for Canadian neutral citations (e.g. 2024 NLSC 12) and common report series, then look each one up through the CanLII API and compare the returned case name with the one in the draft.
  • Idea 2LLM extracts, code verifiesUse a language model only to list every citation and the proposition it is cited for, then check existence with a deterministic lookup so the model never decides whether a case is real.

Specific to this problem

Talk to first

  • CanLII ↗

    runs the free legal database the court names and issues API keys

  • Law Society of Newfoundland and Labrador

    regulates lawyers in NL and can say what checking lawyers are expected to do

Starter prompt

Built from this problem's brief. Paste it into Claude or any AI assistant.

Help me design and build a first working prototype for this problem in Newfoundland and Labrador.

Problem: People filing in NL courts can't quickly check that every case they cite actually exists.
Who it hurts: Lawyers and self-represented people who use AI tools to help draft court filings, and the court staff who read them.
The moment it hurts: Just before a filing goes in, when someone has to confirm each cited case and statute by hand.
What people do today: The NL Supreme Court asks people to verify citations against reliable legal databases such as CanLII. That is done one citation at a time.
Why it is still unsolved: Canadian citations come in several formats (neutral citations, report series), and the duty to check sits with the person filing, who may have no legal research tools.
What solved looks like: A free web tool where someone pastes a draft and gets back every case and statute citation in it, each marked as found on CanLII (with a link), not found, or found with a different name or year. Tested on 20 real Canadian decisions with planted fake citations, it catches every planted fake and wrongly flags fewer than 1 in 10 real citations.
Where: Newfoundland and Labrador

First, list what you would need to learn from the people affected before building, as interview questions.
Then propose the simplest version that would help, the screens or messages it needs, and the data it would use.
Keep it cheap to run, accessible, and respectful of people's privacy. Flag anything you are unsure about instead of guessing.
  1. Understand it

    Talk to the people who live it before you build.

    “Truth is one; the wise speak of it in many ways.”

    एकं सद्विप्रा बहुधा वदन्ति

    ekaṃ sad viprā bahudhā vadanti

    Ṛgveda · 1.164.46
  2. Prototype with AI

    Get a working version in front of people fast.

    Using AI is not cheating. Every tool carries the work of the people who made it, and using it well is how knowledge has always moved forward. The thinking, the care for the people who live the problem, and the judgement are still yours.

    “The Analytical Engine weaves algebraical patterns just as the Jacquard-loom weaves flowers and leaves.”
    Ada Lovelace · Notes on Menabrea's Sketch of the Analytical Engine, Note A (1843)
  3. Find the data

    Open data that fits this problem.

    “The first principle is that you must not fool yourself, and you are the easiest person to fool.”
    Richard Feynman · Cargo Cult Science, Caltech commencement address (1974)
  4. Build it

    Tools matched to what this problem needs.

    “Act, keeping in view the welfare of the world.”

    लोकसंग्रहमेवापि सम्पश्यन्कर्तुमर्हसि

    lokasaṅgraham evāpi sampaśyan kartum arhasi

    Bhagavad Gītā · 3.20
  5. Launch and get support

    Ship it responsibly, and find help and funding.

    “Whatever a great person does, others follow.”

    यद्यदाचरति श्रेष्ठस्तत्तदेवेतरो जनः

    yad yad ācarati śreṣṭhas tat tad evetaro janaḥ

    Bhagavad Gītā · 3.21

Kit written for this problem from its brief; every link was checked. Links go to each provider's own site. Zetema is not paid for any listing.