AKSHĀYVĀT
अक्षयवट · a living tree of knowledge
Brief it like a junior. Review it like a partner.
The firm's own agent for working matters. It reads the record, builds chronologies, drafts with citations to exhibit and page, and commits nothing without your signature.
Private beta with select Indian firms · 2026
When does limitation start running on the s.138 complaint here, from the cheque return memo or from the notice?
Limitation does not run from the cheque return. The cause of action arises 15 days after the drawer receives your statutory demand notice; the one-month window to file runs from there.
On this matter: memo dated 2 Mar 2026 [C3], demand notice 12 Mar 2026 [C5]. Deemed service + 15 days puts cause of action at roughly 30 Mar; the complaint is due by 30 Apr 2026.
Good. Build me the chronology for the evidence bundle from the same documents.
That is the affidavit-chronology skill: every dated event, cited to the exhibit and page it comes from, for your review.
The assistant is docked on every screen. ⌘K from anywhere: a slash command, or a plain sentence.
It moves into the Microsoft 365 estate you already run.
Delegate the work. Keep the decision.
AI does information and technique. Lawyers hold judgment and wisdom. The loop is built so that division cannot be skipped.
Reads the record
Every document read and understood once. Chronologies, evidence reports, and answers, each cited to the exhibit and page it came from. No citation, no entry.
Drafts in the firm's manner
Show it how the firm drafts an instrument once; that becomes a skill anyone can run. Notices, agreements, affidavits, each in your own house style.
Nothing counts until you sign
No output lives in a disappearing chat. Work arrives in the review queue with its sources attached, and joins the matter only on your signoff.
It watches the clock
Limitation and filing deadlines tracked and flagged at 90, 30, 7 and 1 day. Hearings and orders pulled nightly from eCourts, straight to the calendar.
A permanent record
Every action, every approval, written to a record that cannot be quietly edited, and that checks itself every night. Who did what, and when, is never in doubt.
AI you control
Privileged work can run on in-house models that never leave the building. Reach for frontier models only when the work demands it. Your choice of engine.
The firm's own AI, not a seat on someone else's.
The general-purpose AI harnesses are built for software teams and billed by the seat. Akshāyvāt runs on the firm's own AI budget, routes everyday work to efficient models, and keeps the costly ones for when reach truly matters. The firm sets each member's profile and stays in control of the spend. Built for a law firm's economics, not a software company's.
A tree of knowledge that grows with your firm.
We believe no one should have to sacrifice their privacy or their privilege to benefit from AI. Akshāyvāt grows a tree of the firm's knowledge that compounds like a banyan: it sustains the firm, and helps its lawyers practise more efficiently. Our commitment is simple: an institutional, self-learning tree of knowledge, that is the firm's to keep.
Be among the first firms to practise this way.
A private beta is opening to a small number of Indian firms. Ask for a walkthrough, or leave your details for early access.