OUR STORY
Built for the work.
From the inside out.
This is the story of Argent.
Why and how it was built, by the field, for the field.
WHERE THIS STARTED
Forensic psychological evaluation is among the most consequential work a psychologist can do. The opinions we form shape sentencing decisions, determine the potential loss of civil freedoms or bodily autonomy, resolve custody of children, etc. The stakes are not theoretical.
And yet the infrastructure supporting that work has always been limited by the capacity of record volume demands. Not in the clinical training, not in the professional standards, not in the rigor of the methodology — but in the tools. Specifically, the lack of tools for managing records. It’s only been in the last decade that records reliably came out of bankers boxes, and yet our review process is the same as it’s always been.
A single forensic evaluation can generate thousands of pages of collateral documentation. Medical records, psychiatric hospitalizations, school records, law enforcement reports, prior evaluations, depositions. Every page reviewed. There’s no getting around that – Argent users still read every word. But here’s where Argent helps. First, in lightening the load we carry, where every fact is held in our memory or notes across dozens of documents. Second, in the time spent organizing, searching, or collating. Looking back through for that fact you know you read but forgot to note the page number. Pulling out every painstaking arrest to list into an organized chronology. Manually listing the name and dates of every record set reviewed. Every hour spent on those administrative and organizational tasks is an hour not spent on the analysis, the opinion, the work that actually requires a trained forensic expert.
That has always been the tradeoff. It has also always been a professional risk. At sufficient volume, the limits of what any expert can hold across a record set become real — not because forensic psychologists are careless, but because the cognitive demands of comprehensive manual review at scale are genuinely significant. Critical events get underweighted. Patterns stay buried. The foundation of the opinion is only as complete as the review that built it.
Across your careers, those were the conditions of practice. The records arrived. You read them. You held what you found. You formed your opinion. You testified. You went home and started again.
THE IDEA
The arrival of capable AI.
The emergence of capable AI document processing tools created an obvious opportunity for forensic record review. The technology existed to process large document sets, extract structured information, and answer natural language questions about document contents. The application was clear.
What also became clear, quickly, was that no one was building it for forensic psychologists. The tools that existed were built for lawyers, for physicians, for general knowledge workers — each of them useful in their context and wrong for ours. Forensic psychology has specific requirements that generic tools don't address: the citation rigor that adversarial cross-examination demands, the ethical architecture that professional standards require, the workflow integration that makes a tool usable rather than burdensome, the domain knowledge that makes outputs clinically meaningful rather than plausible-sounding.
A forensic psychologist evaluating a generic AI document tool faces a version of the same problem they face when evaluating any instrument not designed for forensic use: the tool might produce results, but it may not produce results you can defend. And defensibility is not a secondary concern in forensic practice. It is a primary one.
So what began as a search for an existing solution became, over eighteen months, the construction of a new one. Not a technology company's interpretation of what forensic psychologists need — but a forensic psychologist's own answer to the problem they had been living with for fifteen years.
THE BUILD
The standard was this: would a forensic psychologist stake their professional reputation on this output?
That question shaped every aspect of how Argent was built. Not the question of whether the AI could produce a result — it can, almost always — but whether the result is one a forensic expert can verify, defend, and stand behind in an adversarial proceeding.
The citation architecture exists because of that standard. A summary without citations is professionally unusable — it requires the expert to take the AI's word for facts that may be challenged under cross-examination. The black box problem can’t be the finish line, and any workable solution needed to shine a light on the output, not the process. A summary with page-level citations is a different instrument entirely: it is a starting point for expert verification, not a replacement for it.
The AI boundary exists because of that standard. Argent does not form opinions, interpret clinical findings, or draw forensic conclusions — because those functions belong to the expert, and any tool that approximates them is not helping the forensic psychologist do their job.
The ethics framework — the APA alignment, the ABFP cooperation framework, the informed consent templates, the report disclosure language — all of it exists because of that standard. A tool built for forensic use that doesn't address the professional obligations of forensic practice is not built for forensic use at all.
The HIPAA architecture, the BAA, the no-training commitment — all of it, too, exists because of that standard. A tool that handles forensic case files containing Protected Health Information and doesn't treat that data with the gravity it deserves is not a professional tool. It is a liability.
Argent was built to pass a test that plays out every day in Courtrooms everywhere: could a forensic psychologist use it, disclose it, defend it, and feel proud that their work product improved from it?
The standard was not whether the AI could produce a result. It was whether a forensic psychologist could stake their professional reputation on it. Every design decision in Argent was made against that test.
FOUNDER
SEBASTIAN RILEN, PSY.D., LP, ABPP
BOARD CERTIFIED FORENSIC PSYCHOLOGIST
Argent Forensics was founded by Sebastian Rilen, Psy.D., LP, ABPP. Dr. Rilen has fifteen years of practice experience in criminal, civil, and family forensic evaluation contexts. He is a board member for the American Board of Forensic Psychology and the Minnesota Board of Psychology. Dr. Rilen presents nationally on Artificial Intelligence in forensic psychology, and authored peer-reviewed literature on the integration of AI tools in forensic psychological practice.
SUPPORT
Need Assistance?
How do I get help if something goes wrong?
Use the **Report an Issue** form — accessible from the footer on every page of the site and from within the platform post-login. Describe the issue in as much detail as possible, including what you expected to happen and what actually occurred. You can expect a response within one business day. Issues involving active cases are prioritized.
What if I identify an inaccuracy in an AI-generated output?
Report it through the **Report an Issue** form. Every reported output error is reviewed. You will receive a direct response explaining what we found and what we are doing about it. Transparent error reporting is a core commitment of the platform — we take inaccuracies seriously and use reported errors to improve the platform.
How do I contact Argent directly?
Support inquiries can be submitted through the Support Form, below or by emailing support@argentforensics.com. For general inquiries, hello@argentforensics.com can be diverted to the founder directly as well.
Get In Touch
We take every report seriously — especially concerns about AI output accuracy. You can expect a response within one business day. If your issue involves an active case, please note that in your description and we will prioritize accordingly.
