Skip to main content

The person behind the public work

SoftwareTeacherPublicBuilder

Hi, I'm Jai Bhagat. I help people choose the first AI project that fits their work, test it, and make the next decision without depending on me.

Book AI guidance · $125Why I am building this
Start with the personShow the proof
10+ yearssoftware and teaching
4.9 / 5.0instructor rating
HashiCorpproduction engineering
Parsonsadult instruction

The honest version

This is the
in-between
period.

For years, I taught economic empowerment work for free. I am now building a public effort that can support more people without depending only on one to one sessions.

Dharmic Data is not a nonprofit today. Between now and June 2027, paid sessions support the time to publish working demos, document what succeeds and fails, and learn what a future collective should become.

I have spent more than ten years moving between software and teaching. I shipped production software at HashiCorp, taught JavaScript at Parsons, and helped Queens founders and operators use AI in work they already care about.

Jai speaking about making a useful project with AIDogged Pursuits photo strip showing Jai teaching and breathing

What makes the work different

Choose the job
before the tool.

I pursue AI when it gives a person more control over useful work. I pass when a simpler system is safer or clearer.

  1. 01

    Start with real work

    Name the recurring problem before choosing a model.

  2. 02

    Use real constraints

    Test against the files, repository, data, time, and machine that actually matter.

  3. 03

    Set the boundary

    Decide what the model may read and which choices always need review.

  4. 04

    Make one small test

    Measure one useful task before expanding the system.

  5. 05

    Publish the proof

    Show sources, traces, evaluations, limits, and failed attempts.

  6. 06

    Know when to pass

    Use ordinary code or a person when AI is not the right tool.

Shakti showing a New York City building recordLIVE PUBLIC CASE STUDY

Proof, not a promise

Shakti shows the method in public.

Its public housing lookup uses ordinary code, not AI. It shows City sources, protects unit details, and makes the address match visible. The optional local AI explains a treated packet but cannot change the record or next step.

Try ShaktiInspect the evidence

Now through June 2027

Paid guidance
funds the
public work.

Paid sessions fund the time to teach, build, publish, and test what a future collective should offer.

  1. NowRent the calendar

    One-to-one guidance pays for time to teach, build, and publish.

  2. NextTest small learning rooms

    Private cohorts, short courses, and a Dogged Pursuits practice room may run with clear, modest promises. They are not for sale yet.

  3. June 2027Seed the collective

    Use the evidence from this period to shape a nonprofit collective for public demos, practical teaching, and shared methods.

The offer today

Bring one real problem.
Leave with a test you can run.

In 60 minutes, we define a use case, choose a tool, set data boundaries, and write a 30 day test plan.

Book AI guidance · $125

Keep exploring

See the work,
then decide.

Back to the homepageRead the build stories