Work on knowledge-first practice, in the open

What you would work on

The internship is built around the three pillars of this site — research, standards, adoption — plus the public-facing craft of the site itself. Depending on your background and what is live at the time, you might work on:

• Research pieces — developing and refining short, self-contained position pieces on knowledge as a first-class layer in agentic systems: why it matters, how it behaves in enterprise and personal contexts, why context windows are a knowledge-delivery problem, where token cost and grounding really come from.
• Standards and conventions — shaping and documenting the open conventions for representation, lifecycle, grounding, the agent interface, and discoverability, and testing them against real settings.
• Adoption practice — helping shape the maturity evaluation, lifecycle adoption, audits and benchmarks, advisory, and training practice that turns the ideas into something a team can run.
• The public site — building and refining the site itself: the pages, the chrome, the blog, the free resources, the agent surface, and the discoverability layer. This is a small, flat site by design; that makes it a good place to learn the craft without being buried in a large codebase.
• Community work — helping run the public side of the community: the Bengaluru JavaScript Meetup, the discussion around the ideas, and the shared learning that surrounds the work.

The work is not a single project. It is a slice of the open effort, sized to what an intern can carry and what the site needs at the time.

Who this is for

This internship is for someone who wants to work on the ideas behind agentic AI — not just the models, but the knowledge layer that makes them dependable — and who wants that work to be public, open, and not-for-profit.

It is a good fit if you are interested in one or more of:

• Knowledge management and knowledge representation, in a setting where the ideas are being put to work, not just studied.
• Agentic AI, LLMs, and personal assistants — and the failure modes that come from weak knowledge rather than weak models.
• Writing and thinking in public — short, self-contained pieces that open with the direct answer and then go into enough substance to be useful on their own.
• Web craft — building a small, flat, light-themed site in HTML and CSS, with a real audience and a real voice.
• Community — running and taking part in public discussion, including a meetup with a real audience.

The internship is in Bengaluru. It can be on-site, hybrid, or remote depending on the person and the timing; the default is hybrid, with the work visible in the open either way.

What you get out of it

The internship is not a line on a resume. It is a chance to do real work on real ideas, in public, with a real audience. Concretely:

• Your name on work that is published on the public site — research pieces, standards, adoption practice, blog posts, and the site itself.
• Direct exposure to the failure modes and the practice of knowledge-first agentic AI, from someone who has built and scaled an AI Centre of Excellence and is writing a PhD on multi-agent communication.
• A public body of work you can point to after the internship — not an internal report, but something on the web, with your name on it.
• A real reference from a small, public effort — which is often more useful than a reference from a large, anonymous one.

The internship is paid where possible and unpaid where it cannot be; the standard is that the work is real and the credit is public. If that does not fit your situation, say so up front and we will talk about what can be done.

How to apply

To apply, write to janardan.revuru@gmail.com with:

• A short note on what interests you in the work — not a generic cover letter, but a specific take on which of the three pillars or the site craft draws you in.
• What you have already done that is relevant — a project, a paper, a blog, a meetup, a piece of open work, anything public that shows how you think.
• What you are looking for — a season, a topic, a location, a paid or unpaid arrangement, anything that matters to you and that we should know before we talk.

The bar is not credentials. The bar is whether the work is a real fit for the person and the person is a real fit for the work. The public site is small enough, and the ideas are sharp enough, that the right person can do real work fast.