Questions worth asking

The ones firms actually ask, answered plainly. If something here is not clear enough to take to your risk team, tell us and we will fix the answer.

Leonard

Who is Leonard?

Leonard is Legal Engine's submissions interviewer. He is not a chatbot bolted onto a form: he is a voice agent built for one job, which is to sit with the lawyer who did the work and ask what a submissions expert would ask, in the order they would ask it.

He does not write your submission for you. He captures what the lawyer says, fills your template in front of them, and hands back a Word document they have already checked. Everything in that document is traceable to something the lawyer actually said.

Holding copy, pending Andy's wording.

How it works

What actually happens when a lawyer uses Leonard?

They follow a link, sign in with a one-time code, and talk for a few minutes. Leonard asks the questions a submissions expert would ask, in the order they would ask them. The submission template fills in on screen as they speak. At the end they check it, correct anything that is wrong, and download the Word file.

How long does it take?

Around five minutes for a straightforward matter. The conversation adapts: a matter with more to say takes longer, and one where the lawyer says "skip that" moves on.

Can lawyers change what Leonard captured?

Yes, any field, by clicking on it. During the conversation or after it. A value a lawyer has edited is never overwritten by anything Leonard hears later, which is deliberate: nothing is more corrosive to trust than watching your correction disappear.

What if Leonard mishears something?

You see it, because the field fills in front of you rather than after the call. Where Leonard is not confident, the value is shown greyed rather than asserted as fact. And a value only appears once the words that produced it have been said, so it never runs ahead of the conversation.

Does Leonard submit it for us?

No, and deliberately. It produces the finished document. Whether and when that goes to Chambers stays a human decision.

Which directories does this cover?

Chambers & Partners as standard. Legal 500 is an add-on that uses the same conversation and the same generation flow.

What you get

What comes out at the end?

A Word document on your current template, with the confidentiality markings and shading applied from what the lawyer actually said. It is generated to match the document your BD team produces today, so nothing downstream has to change.

How does it know whether a matter is confidential?

It asks, and it asks twice, because they are two different questions. Whether the firm may reference the matter externally is asked separately from whether the client may be named. A client agreeing to be named does not mean the matter can be used, and treating them as one question is how the wrong thing ends up published.

If the answer is unclear or was never given, the document is marked confidential. Over-marking is recoverable; publishing a client name that should not have been is not.

What is the data export add-on?

Every conversation is captured as structured fields rather than filed as a transcript. The export is all of it in one file, in a schema aligned to the submission template, at the end of the season.

What can we do with the export?

Load it into pitch management software, run your own model across it, or use it for credentials and marketing outside directory season. It is your data in a standard shape, not a report we generate for you.

What format does the export come in?

CSV, XLSX or JSON. Whichever your systems want.

Security and data handling

Where is our data processed?

Compute is pinned to the UK and EU: the application runs in London and the language models are called in EU regions. Speech processing uses EU residency endpoints.

We describe this as the region being pinned rather than as a residency guarantee, because those are different claims and only the first is one we control end to end. If your risk team needs the second, ask us and we will give you the full picture of every subprocessor rather than a slogan.

Is our data used to train models?

No. Conversations are not used to train any model, ours or a supplier's.

How do lawyers sign in?

A one-time code sent to their firm email address. There is no shared password and no account for anyone to create. Access is scoped to your firm.

What happens to our data when the contract ends?

You get the export if you have taken that add-on, and the conversation data is deleted. We will put the specifics in the contract rather than in a FAQ, because that is where it belongs.

Are you ISO 27001 certified?

Yes. Legal Engine is certified to ISO 27001.

All data is hosted and processed in the EU, and nothing is retained once the contract ends. If your risk team needs the certificate or the scope statement for a procurement questionnaire, ask and we will send them.

Who do we contact with security questions?

Email us and it goes to a person, not a queue. We would rather have the conversation before you buy than during a review.

Commercials and rollout

How is pricing structured?

A flat fee per submission season, by firm size. Not per seat, not per submission. Directory season is a fixed piece of work with a fixed deadline, so it is priced that way.

Is that a genuine flat fee?

Yes. Unlimited conversations for every eligible lawyer at your firm, for the season. If it were metered, the people who most need to use it would be the ones told to be sparing with it.

What currency?

Pounds sterling, excluding VAT.

How do we roll this out to our lawyers?

You get a firm-specific link to circulate. There is no software to install, no portal to learn and no training session to sit through: if a lawyer can join a video call, they can use this.

Still have a question?

Ask it. If the answer is not good enough we would rather hear that now.

Talk to us