Good ideas no longer need a sponsor before they can exist: vibecode.law co-founder Alex Baker

The Indian LegalTech Network (ILTN) and vibecode.law have launched the Vibeathon 2026, billed as India’s first “vibe-coding” hackathon for lawyers.

The initiative gives lawyers, law students and other legal professionals three weeks to identify problems in legal practice and build working solutions using AI-assisted coding platforms. Participants do not need prior coding experience, with the organisers seeking to demonstrate how AI can lower the barriers to building software for legal practice.

vibecode.law, launched earlier this year by Chris Bridges of Tacit Legal, Matt Pollins of Lupl and Alex Baker of Legal Tech Collective, began as a weekend project between its three founders.

Baker noted that the platform has since seen hundreds of lawyers from around the world build a wide variety of applications and share them with others exploring legal technology.

Baker further spoke about what prompted the creation of vibecode.law, where lawyers should begin if they have never coded before, the limits of AI-built prototypes and what vibe coding could mean for innovation within law firms.

Edited excerpts follow.

Question: vibecode.law started as a weekend project between the three of you. What did you think would happen, and what actually happened?

Alex Baker (AB): We thought about twenty people from our network would put projects up and that would be a good start. But we've had 100s of lawyers building a huge variety of applications from all over the world. The platform ended up being more about making the activity of lawyers developing solutions visible to others who are interested in exploring.

When you can see that an associate in another part of the world built a version of the thing you have been thinking about, the mindset moves from "could this work" to "how did they do it".

Question: Where should an Indian lawyer starting from zero begin?

AB: Not with a tool. Start with a problem you have every week, that irritates you, and that does not involve client data. If you have never built anything, start with something where you describe what you want in plain English and immediately see a working screen.

Lovable and Google AI Studio both do that, and both have free tiers that are more than enough to find out whether you enjoy it. 

Your first build should use invented data, not anonymised client data. Prove to yourself you can make something that works, then have the harder conversation about pointing it at real data. 

Question: What's the realistic thing a lawyer can build in a weekend, and what still needs a developer?

AB: In a weekend, with no coding background, you can build something that works properly for one person. That is not a small thing.

The pattern that works is single user, internal facing, no confidential data, and a task you currently do by hand. Checklists that flag missing provisions. Something that turns a messy email chain into a chronology. A generator for a document you produce in the same shape twenty times a year. A tracker for regulatory updates in your area.

What still needs a developer is everything that follows from other people relying on it: client data at scale, anything talking to your document management system, user accounts and permissions.

Question: What's the path from personal prototype to something a firm actually uses, and where does it usually break down?

AB: You build it for yourself. You use it for a few weeks and it survives contact with reality. A colleague asks for it or you build a business case around how much it could be used. Now you have evidence, which firms find difficult to argue against.

If you have no development experience, taking something from prototype into production is risky. How can you tell if it is secure? Even if you managed to get it right, if you and others start to rely on it, what happens when it breaks? That doesn't mean there isn't value in the process - the prototype is a precise specification of what you want to develop.

Question: Indian law firms are more cost-sensitive and more hierarchical than UK or US firms, and juniors don't typically get to ship anything. Does vibe coding change that, or does it just give partners another thing to approve?

AB: Vibecoding collapses the cost of trying something but it won't change the culture of any firm regardless of the part of the world you operate in. What it does change is what a junior can bring to the table. An associate with an idea used to need a partner to sponsor it before anyone could assess whether it was any good, because assessing it meant a budget and a developer. 

Now that associate can arrive with a working prototype and much more evidence around its potential and the impact it could make. The hierarchy will remain, but now good ideas no longer need a sponsor before they can exist.

Firms that encourage their juniors, who are typically more technically savvy employees to experiment are going to find new ways of delivering value to clients.