Why I Built BriefVision: Legal Evidence Analysis Software Born from Evidence Overload
Updated: Aug 4
How preparing for a rare and complex criminal trial led me to build a secure, AI-assisted workspace for reviewing documents and digital evidence.
CONTENT NOTE — This post discusses bereavement, suicide, serious criminal allegations and a criminal trial. If you are affected by any of these issues, support information appears at the end.
BriefVision is now legal evidence analysis software for complex cases, but it did not begin with a product roadmap. It began with a problem I could not afford to leave unsolved.
In December 2024, I was charged with manslaughter, two counts of rape, and coercive or controlling behaviour in connection with the death by suicide of my late wife, Tarryn Baird, in 2017. I maintained my innocence throughout. On 22 April 2026, after more than 40 hours of deliberation, a jury at Winchester Crown Court found me not guilty on every count.
Domestic abuse is real and serious, and people who report it must be supported. This post is not an attempt to diminish that truth or to retell every part of a painful case. It is an account of the practical problem that confronted my legal team and me, and how solving that problem became BriefVision.
When the evidence becomes the problem
Once disclosure began to arrive, the scale of the task became clear. There were thousands of pages of exhibits, witness statements and unused material. Some documents were searchable; many were not. Dates appeared in different formats. Relevant facts were scattered across PDFs, spreadsheets and separate evidence bundles.
Later, the prosecution served a download from Tarryn’s iPhone as a spreadsheet containing more than 392,000 rows. It included messages, emails, location information and other device data. Even on a powerful laptop, searching and filtering it was slow. Reading it as a flat spreadsheet also stripped away the context that makes digital evidence understandable: conversations did not look like conversations, email bodies could appear as raw HTML, and coordinates were simply numbers.
I knew my legal team could not spend unlimited time fighting the format before they could examine the substance. So I began building small tools to make the evidence workable.
From small scripts to legal evidence analysis software
The first steps were simple and practical. I split very large document bundles into manageable files. I ran image-only documents through optical character recognition so they could be searched. I wrote code to recognise dates such as “1 May 2016”, “01/05/2016” and “01.05.2016” and store them in a consistent form.
Next came indexing. Instead of opening one file after another and repeating the same searches, I created a searchable store of the case material. I then added an AI-assisted question-and-answer layer that could work from the case data and point back to the material supporting an answer.
When the phone spreadsheet arrived, I loaded its rows into a database and built specialist views around the data. What had begun as a collection of utilities was becoming a single workspace for legal evidence review.
Making digital evidence readable
Digital evidence should retain its source and metadata, but it also needs to be presented in a form a human reviewer can follow. BriefVision developed around that principle.
Messages are grouped into conversation threads and can be filtered by device, source application, date, attachments and search terms.
Emails are rendered as readable messages rather than raw code, with sender, recipient, mailbox, folder, attachments and other metadata kept alongside the content.
Locations can be viewed on maps and examined by date, source, device, coordinates and available accuracy information.
Calls, contacts and calendars have dedicated views that preserve their evidential context while making them easier to search and compare.
Timelines bring dates and events from different sources into a chronology that can be filtered and tested.
Today, the Digital Sources module brings these specialist views together with the wider case record, so documents and device data can be reviewed in context rather than in isolation.
AI assistance that stays anchored to the evidence
AI can be useful in a complex case, but only if the reviewer can return to the underlying material. BriefVision is designed to search, summarise and answer questions using the evidence in the workspace, while pointing the user back to the source that should be checked.
That distinction matters. The software does not decide what a fact proves, replace disclosure obligations or substitute for legal judgement. It helps reviewers find potentially relevant material, understand connections and spend more of their time on analysis and strategy.
What it changed for my defence
The value of the system was not any single dramatic search. It was the ability to test a date or allegation across different sources without losing the chain back to the original evidence. A statement could be compared with travel records, emails, messages, location data and other documents in a common chronology.
That made it easier for the legal team to identify contemporaneous records, examine inconsistencies and prepare the case. The advocacy, legal analysis and courtroom work belonged to my solicitors and counsel; the verdict belonged to the jury. BriefVision was a tool that helped us work through an extraordinary volume of material with more clarity.
I remain deeply grateful to Rishi Verma of ABV Solicitors, Katy Thorne KC and Omran Belhadi, whose work is summarised in the Doughty Street Chambers case note. I am equally grateful to my wife Bea, my family and my friends for standing beside me throughout.
Why BriefVision exists now
Most legal matters will never resemble my case, but evidence overload is not unusual. Criminal, family, civil, regulatory and internal investigations increasingly involve a mixture of documents, messages, email accounts, mobile phone extractions, location records and large spreadsheet chronologies.
BriefVision exists to make that material easier to organise, search and understand in one secure workspace. It is hosted on Microsoft Azure, uses controlled access and retention, and does not use case data to train models. The aim is straightforward: help legal teams move from manual sorting to focused review without losing sight of the source evidence.
I built the first version because I needed it. I am continuing to build it because many other legal teams face the same underlying problem: the important fact is often present, but buried in a volume or format that makes it unnecessarily hard to find.
To explore the product, see BriefVision’s legal review features and its Digital Sources workspace, or book a focused demonstration.
Further context
SUPPORT — If you are struggling to cope, Samaritans can be contacted free at any time in the UK and Ireland on 116 123.
Comments