The questions we get asked, in plain English.

The exports themselves are documents, admissibility in family proceedings is governed by the Family Procedure Rules and the discretion of the court. What we provide is a chain of custody (SHA-256 hash chain per page, audit log per case) that helps the document satisfy authenticity challenges. Whether a particular document is admitted in a particular hearing is for the court.

Hearsay rules apply differently in family proceedings (especially in children matters) than in civil. We don't determine the hearsay status of evidence, that's a matter for counsel. We do preserve the original form of the communication and its metadata so the court can assess for itself.

The SHA-256 hash on every page of the export lets the other side independently verify that the document we produced matches the source. We also publish the hash of the input file in the audit log so they can verify nothing was modified between intake and export.

Yes for FPR. CPR-style disclosure lists are on the roadmap; ask if you need it now.

No. Customer evidence is never used to train models. Where third-party model endpoints are used (e.g. for transcription), we use enterprise endpoints that contractually prohibit retention and training use.

UK-based infrastructure (London region). No transit outside the UK except for your own export downloads.

Case data is retained for your 12-month access period. After that, data is preserved in read-only mode for 30 days then automatically deleted. You can trigger immediate deletion at any time and we'll return a signed deletion tombstone.

Yes. Available on request via the contact form.

You are responsible for the lawful basis on which you're processing third-party data. The most common bases in family proceedings are "necessary for the establishment, exercise or defence of legal claims" under Article 9(2)(f) of the UK GDPR and the corresponding DPA 2018 Schedule 1 condition. We don't audit your basis but we expect you to have one.

Accuracy varies by data quality. On well-formed WhatsApp exports the timeline, author breakdown and counts (apologies, profanity, who opens hostile exchanges) are essentially exact. Tone scoring is an indicator, not a finding. Pattern detection (late handovers, missed payments) is highly accurate when the underlying messages are present and dated. The tool reports what the record contains; it does not decide who is right.

Every AI-surfaced item is footnoted to a source quote with its timestamp. If we're wrong, you'll see we're wrong, and you can exclude it. Because everything is grounded to the underlying message, nothing is asserted that the record can't support, and the platform never draws a conclusion about who is at fault.

No. We replace solicitor reading-time. Advice, drafting, negotiation, advocacy, those are still your solicitor's job. We just stop you paying £375/hr to read your own chat history.

Where engaged by a solicitor as part of a specific matter, we operate within the solicitor's evidence-handling workflow and the privilege attaching to that workflow. Where engaged direct by an individual, we are not subject to LPP, but our staff are bound by NDAs.

If you have a solicitor, it's usually better for them to engage us. That way the work product is covered by LPP and the solicitor signs off on what's exported.

One-off charge, single chat upload, 2-page report. No subscription, no auto-renew. Designed as a try-us before committing.

For consumers it's a one-off payment per case, with 12 months of access. No auto-renew. You keep read-only access to existing exports forever. Only firms running many cases pay monthly.

14 days from purchase on Single Case and Complete Case. Taster: refundable if export fails. Court Bundle: refundable up to the point the bundle is generated.

No. LegallyHeard is fully automated, no human reads, reviews or supervises your conversations. Your evidence is processed by the pipeline and only you (and whoever you choose to share the export with) ever see it. This keeps it private and keeps the platform neutral.

A fully automated, court-ready bundle: pagination throughout, an exhibit list, a chronology, a source manifest, a statement-cross-reference document and a SHA-256 hash-chain so integrity can be verified. Complex multi-source cases are quoted from £1,750–£3,500. No human time is billed, you're paying for the bundle the pipeline produces.

We are pre-launch. Early access is being granted to family-law firms and select litigants in person. Sign up via the contact form to join the waitlist.

Be upfront about gaps: the in-product audit-log UI, the CPR-style disclosure-list export, and the multi-seat firm dashboard are on the roadmap. Ask for specifics if it matters for your decision.