iLawBot handles the research, drafting, and groundwork so you stay focused on the work that matters. Citation-backed answers grounded over your own documents and Indian law — built for advocates, firms, and students.

DPDP-aligned
Data-protection by design
BCI-conscious
No prohibited AI uses
In-India inference
Privileged work stays in-region
Verifiable citations
Every claim is openable
One connected workspace
From research and drafting to OCR and case management — iLawBot brings your documents, the law, and your reasoning into a single, verifiable system.
Traverse a 3D citation graph built over Indian case law and your own library. Treatment signals — followed, distinguished, overruled — are clearly labelled as AI-suggested and as-of a date, never asserted as settled good law.

A conversational research assistant answers over your documents and Indian law. Every claim carries an [id] marker you can open and verify, with an optional web-search tool that is clearly flagged as unverified.

Generate structured petitions, agreements, notices, and briefs — or refine existing drafts. Output is always advisory and reviewable, with the law reasoned as-of the relevant date (BNS/BNSS/BSA from 1 Jul 2024).

Upload pleadings, contracts, judgments, and handwritten land records. Each file is OCR'd, chunked, embedded, and indexed — with low-confidence regions routed to a human-verify queue and originals always retained.

Keep clients, matters, and their relationships in one deterministic workspace, with a hearing calendar and cause-list view plus limitation-deadline awareness so critical dates surface before they bite.

Built for Indian practice

Coordinate matters, evidence, and authorities across the bench.

Build evidence timelines with BNS-2023-aware reasoning.

Review contracts and surface risk with an AI assistant.

Research, draft, and manage cases from a single workspace.
How it works
Upload pleadings, contracts, and judgments — even handwritten records. Each file is OCR'd, chunked, embedded, and indexed, securely and tenant-scoped.
Research over your library and Indian law, draft documents, and traverse the citation graph — all in one connected workspace.
Open every citation, check the grounding report, and apply human review. AI output is advisory — nothing files itself.
Trust isn't a feature you bolt on. These constraints are load-bearing — enforced in the product and asserted by our tests.
Answers cite only retrieved, verifiable spans. Good-law signals are labelled AI-suggested and unverified unless from a licensed citator.
Every AI output is advisory. Low-confidence OCR/HTR regions route to a human-verify queue; originals are always retained.
A privilege classifier runs first. Privileged content is pinned to the in-region path and never egresses to external providers.
Claims, citations, scores, and human actions are recorded immutably — a complete, tamper-evident trail of every decision.
We reason about which law applied as of the offence/transaction date (BNS/BNSS/BSA from 1 Jul 2024) — not a naive 1:1 map.
No adjudication, sentencing, bail/recidivism scoring, judge analytics, or outcome prediction. Ever.
Start free and upgrade when you're ready. Pick a plan, sign up, and you're in — no gaps, no surprises.
Yes. It is India-native: BNS/BNSS/BSA-aware temporal reasoning, Indian states and languages, and a workspace designed around how advocates and firms actually work.
Always. Every answer cites retrieved spans by [id]. Open the source, read the surrounding context, and review the grounding report before relying on anything.
A privilege classifier runs before any model call. Privileged content is pinned to the in-region Workers AI path and is never sent to external providers; the egress fallback is disabled on that branch.
No. The Explore plan is free. Paid plans add capacity and team features; you can start free and upgrade any time.