Why We Don’t Sell Tech: The Case for Tech-Agnostic Review

The legal industry is in the middle of a loud conversation about AI in document review. Vendors are racing to release new platforms, bundle their own tech, and position themselves as one-stop shops. It’s a reasonable business strategy. It’s just not ours.

At Legalpeople, we’ve made a deliberate choice: we don’t sell technology. We partner with every major eDiscovery vendor and work inside whatever platform our clients prefer, from Relativity to Reveal to Everlaw to whatever comes next. That’s what technology-agnostic document review means in practice: our Managed Document Review teams of licensed attorneys and experienced project managers work in the tools you already use.  What we provide is the people. That is the business. That has always been the business. In 2026, as AI takes on more first-pass review, the people left doing the work carry more of the risk. That’s why our focus matters.

A narrative is floating around that AI will eliminate the need for document review attorneys. The reality on the ground looks different. Technology-assisted review (TAR) and generative AI are reducing the volume of documents that reach human reviewers, yes. But the work that remains is harder and carries higher stakes than first-level review ever did.

Review teams today are being asked to:

• Interpret data that AI models can’t reliably process, including audio, video, handwritten notes, multilingual communications that call for foreign language document review, and ambiguous business shorthand 
• Validate AI-filtered outputs and reconcile discrepancies between algorithmic determinations and case team strategy
• Conduct deeper analysis when a production needs to support case development, beyond basic responsiveness calls 
• Make contextual judgments about privilege, work product, and sensitivity that require understanding the case, not just the document

This is sophisticated legal work. It requires reviewers who understand the matter, the strategy, and the consequences of getting it wrong. A bar admission and a pulse do not qualify someone for this job anymore, if they ever did. We saw this on a high-profile Second Request, where the eDiscovery vendor’s proprietary analytics tool routed potentially privileged documents to our reviewers unevenly. Human judgment is what kept that review on track. 

The best AI review tools on the market are impressive. They surface patterns, cluster concepts, and accelerate workflows in ways that would have been hard to imagine five years ago. Our clients use them, and we help our clients use them well. They also make mistakes only a trained reviewer will catch, a problem we covered in AI Hallucinations: Why Legal Work Still Needs Humans. 

But there are categories of work where AI falls short, and these gaps are where the actual legal judgment happens:

A model can tell you a document discusses a contract negotiation. It can’t always tell you whether the tone suggests bad faith, whether a specific phrase is a term of art for this industry, or whether the sender’s history with the recipient changes how the language should be read.

AI doesn’t know what the case is really about. It doesn’t know what the partner wants to prove, what the opposing side is likely to argue, or what a particular document might mean three weeks from now when a deposition question turns on it.

When something goes wrong in a production, someone has to answer for it. That someone is a licensed attorney. No vendor is going to put their algorithm on the stand.

Privilege calls, work product protection, and ethical walls require human attorneys exercising professional judgment. The rules of professional conduct do not have an AI exception.

When we founded Legalpeople in 2002, we made a choice to specialize. We don’t sell platforms, because selling platforms pulls a company’s attention toward software development, licensing models, and tech margins, and away from the thing we care about most: recruiting and retaining the best possible legal talent. Staying technology-agnostic keeps our focus there. 

That focus shows up in the numbers. We’ve maintained 95%+ internal management retention. Our senior project managers have tenure measured in decades, not quarters. Since 2022, we’ve made more than 3,300 placements, and we’ve built the kind of reviewer network where we can staff sophisticated, high-stakes projects with attorneys who have the experience the work requires. That network is how we staffed a multi-city Second Request with 250 attorneys across two cities in less than a week. 

Compare that to the staffing model some vendors use, where tech is the real product and the review team is assembled quickly from whoever is available. That model might have worked when first-level review was largely about volume. It doesn’t work when the remaining human review is the work that matters most. It’s the same reason your eDiscovery technology provider shouldn’t be your contract review team, either. 

You choose the platform. We support it. If you have a preferred vendor relationship, we work inside it. Law firms running their own review platforms can bring us in through law firm eDiscovery unit staffing. If you’re evaluating options, we can speak candidly about how different tools perform in different kinds of matters, because we’ve seen all of them in action and we don’t have a horse in the race.

You define the workflow. We bring the talent to execute it. Our project managers have run thousands of reviews across every major industry and matter type, from litigation and HSR Second Requests to regulatory investigations and internal investigations. They know how to integrate AI-assisted document review workflows with attorney judgment in ways that improve quality as well as speed. 

And when something complicated comes up, we have the people who can catch it: a document the model flagged wrong, a privilege question that needs a second look, or a pattern that might matter for case strategy. 

AI is going to keep getting better, and volume will keep going down. The work left for human reviewers will depend more and more on real legal judgment. 

That’s the environment we built Legalpeople for. We didn’t pivot into it when the market changed. We built Legalpeople for this from the start. 

We don’t sell tech because the people are the point. That’s the case for technology-agnostic document review: as AI handles more of the easy work, the quality of your human team is the thing that determines whether your review holds up. 

That’s what we do, and we do it better than anyone who’s trying to do three other things at the same time.

Looking for a Managed Document Review partner who can work within your preferred tech stack and bring the legal talent your matters require? Let’s talk.