Shreya Vajpei founded the Indian LegalTech Network, starting with a coffee shop meetup in Mumbai. It’s now 500+ members across 15 cities. Before that she led GenAI adoption at one of India’s largest law firms. Named an ILTA Influential Woman in LegalTech 2026, a LinkedIn Top Voice for Law Tech and AI, and among the Top 30 Legal Tech Innovators in APAC.
10 years in law and legal tech, now based in London, where she mostly asks what is the bottom line impact of legal AI.
Session Overview
Paying the Verification Tax in the Age of GenAI
Most of us sort legal work for AI by risk. Low risk to the machine, high risk to the human.
It’s the obvious approach, and it’s the wrong axis. Risk tiering prices the consequence of an error. It says nothing about the cost of finding one.
The full cost of delegating a task is the agent doing it plus somebody checking it, and the checking is what dominates. Tools get most of an output right most of the time, and almost never all of it right. When nothing in the output tells you which part failed, you check all of it. Drafting time collapses. Total time doesn’t – the bottleneck just moves.
This session works through what actually makes an output cheap or expensive to check: whether a failure can be located, how many things have to be right at once, whether ground truth exists, whether an error surfaces downstream anyway, and whether the action can be undone. Then we apply it, decomposing a real workflow so the delegation decision happens at the step rather than the whole task.
Three Tangible Learnings
Learning 1: Understand what verification actually costs, and why it determines whether AI saves time at all.
Learning 2: Learn to sort work by verifiability.
Learning 3: Learn how to decompose a workflow so the delegation decision happens at the step, not the whole task.
There’s a Legal AI Vendor for Everything. That’s the Problem
There’s a legal AI vendor for everything now. Several for things you don’t even do. Most of them work.
That’s the problem.
Everyone agonises over which one to pick. Picking is the easy part.
The thing that kills your deployment isn’t visible at the demo. It shows up eight months later, in the corners of the business nobody walked the vendor through. A tool that solves a real problem you happen to have four times a year. A demo run on clean documents when yours are scanned, misfiled and named “final_v3_ACTUAL”. Pricing that quietly punishes the way your team actually works. Two hundred seats bought, forty ever opened.
None of that is a product failure. But you still see nothing on the bottom line.
None of this is new. Every one of these failures was documented decades ago, by industries that burned the money before we did.
Three Tangible Learnings
Learning 1: Understand where legal AI decisions actually go wrong, and why most of it stays invisible until well after you’ve committed.
Learning 2: Learn what other industries already worked out about technology adoption, and what transfers to a legal AI decision.
Learning 3: Learn where honest, unfiltered information about tools and vendors exists, and how to get to it.
Alternatively, please email sales@eventfulpeople.com for quote, invoice or any queries and we’ll be glad to assist.
Subscribe to our mailing list
"*" indicates required fields
Download Agenda
Complete the fields below in order to access our Agenda.
Download the agenda
Please fill out the form to view the Conference Agenda pdf
Community Snapshot
Complete the fields below in order to access our community snapshots.
Community Snapshot
Complete the fields below in order to access our community snapshots.
Find out more about
Planner’s School
Find out more about Planner’s School
Shreya Vajpei founded the Indian LegalTech Network, starting with a coffee shop meetup in Mumbai. It’s now 500+ members across 15 cities. Before that she led GenAI adoption at one of India’s largest law firms. Named an ILTA Influential Woman in LegalTech 2026, a LinkedIn Top Voice for Law Tech and AI, and among the Top 30 Legal Tech Innovators in APAC.
10 years in law and legal tech, now based in London, where she mostly asks what is the bottom line impact of legal AI.
