Neotas Hunts the Risks That Never Made the Watchlist

The awkward thing about a watchlist is that somebody has to do something watchlist-worthy before appearing on it. A sanctions register can tell a bank what authorities already know. A credit file can describe yesterday's solvency. A supplier questionnaire can reveal what a supplier elects to put in a supplier questionnaire. Neotas, a London company founded in 2017, makes its living in the gap between those tidy records and the untidy world they describe.

Its pitch is straightforward: enter a person or company, pull together sanctions, politically exposed person and adverse-media checks, then keep going. Search corporate ownership, court records, archived pages, social platforms and the deep web. Resolve duplicate names. Map relationships. Classify the findings. Put a citation beside every assertion. If the case looks genuinely peculiar, pass it to an investigator.

That last handoff is the interesting part. Neotas is neither only a software subscription nor only a room full of researchers. It sells a staircase from basic screening to an automated due-diligence report and, finally, to analyst-led enhanced due diligence. Most cases should not require an expensive detective. Some cases should never be trusted to a green tick produced in nine seconds.

The original irritationThe questionnaire fails first

Vipul Mishra came to the problem after working in information security at a London hedge fund. His co-founder, Ian Howard, brought more than three decades in finance and investment. Their starting observation was less glamorous than the current vocabulary of generative AI: due diligence was fragmented. An analyst might search databases in one window, the web in another, save notes elsewhere and assemble a report by hand. Each handoff created delay, inconsistency or a missing piece of evidence.

The first failure was often the source itself. Traditional screening is excellent at matching a known person to a known list. It is less useful when the issue is an undisclosed directorship, a cluster of aliases, a local-language court mention or a trail of business relationships three companies removed from the subject. Self-reported ESG and supplier assessments have the same weakness: they begin with the target's account of itself.

Neotas changed the unit of work from the database hit to the investigated subject. A hit without identity resolution is noise. A red flag without its original source is an argument waiting to happen. A report without a log is fragile when a regulator, investment committee or lawyer asks, reasonably, “How did you reach that conclusion?”

“Neotas searches go deeper than traditional due diligence checks by ‘spidering out’ across the entire internet.”Coller Capital, a Neotas client

What it actually doesA machine searches; a person remains accountable

The automated product starts with a name or entity and any identifying seed data. Software confirms the likely subject, strips duplicate results, then runs sanctions, PEP, adverse-media, corporate-registry and deep-web searches in parallel. The output is a red-amber-green report. Findings are classified rather than dumped into an inbox, each one links back to evidence, and the search history becomes part of a tamper-evident audit trail. The report can be edited before export.

For a routine customer, employee or vendor, that may be enough. For a politically exposed person, a cross-border ownership chain, a high-risk jurisdiction or disturbing media, the same subject can climb into analyst-led work. Neotas offers progressively deeper reviews: a human snapshot, a full investigation, and network mapping. Analysts bring the context software is bad at pretending to possess - whether two similar names are the same person, whether sarcasm is actually menace, whether a holding company is ordinary structuring or a useful place to hide control.

The broader platform adds the machinery around that investigation: vendor onboarding, questionnaires, configurable scoring, remediation tasks, continuous monitoring and APIs. That puts Neotas in several markets at once. It competes with AI research tools such as Xapien and Blackdot, workflow systems such as ProcessUnity and Aravo, data services such as World-Check and Dow Jones, and old-school investigation firms such as Kroll and Control Risks. Neotas also integrates several of those data providers. In risk technology, a rival can be both competitor and ingredient.

Who pays, and whyRisk teams buy fewer tabs and a better receipt

The buyers are compliance, procurement, legal, HR and investment teams. Financial institutions use it for KYC, AML and vendor risk. Private-equity firms examine founders and management before committing capital. Companies screen employees in sensitive roles. Healthcare organisations assess vendors carrying patient, operational or regulatory exposure. Investigators use the platform as their workbench rather than outsourcing every case.

Public client references include Coller Capital, Channel Capital Advisors, IK Partners, VCCP and Catalysis Advisory. The last of these had completed more than 100 management due-diligence cases with Neotas by 2020. Neotas said in 2025 that it had supported more than 400 transactions across 53 jurisdictions in the preceding year. Those figures are more revealing than a giant “data points searched” number: customers did not merely log in; a risk decision sat on the other side.

The company raised £1 million in 2018 from Symvan Capital and Force Over Mass Capital. The money was earmarked for the platform, more analysts, sales and marketing. A subsequent Knowledge Transfer Partnership with the University of Essex, backed by nearly £200,000 from Innovate UK, applied natural-language processing and image analysis to reputational-risk research. That collaboration captures what changed the economics: the web became too large and multilingual for manual collection, while the consequences of a careless match remained too serious to automate blindly.

The bars compare categories for visual orientation, not on a common quantitative scale. Neotas reports absolute coverage of 600B+ archived pages, 1.8B+ court records, 198M+ corporate records and 40,000+ media sources.

The useful imitationCopy the escalation rule, not the data boast

A compliance team does not need Neotas's index to borrow its most practical idea. Divide cases by the decision they could change. Let cheap, consistent screening clear ordinary names. Require an evidence-backed report when signals conflict. Reserve investigators for material ambiguity. Most importantly, define the trigger before the alarming result arrives. “High risk” is not a workflow.

The same applies to AI. Automate collection, deduplication, translation and first-pass classification. Do not automate responsibility. Neotas advertises a model-agnostic design, allowing clients to switch the underlying AI rather than lock a regulated process to one vendor. The more durable asset is therefore the evidence trail, scoring logic and escalation practice around the model.

The team behind the workflow is deliberately mixed. Neotas describes backgrounds in government, law enforcement, fraud prevention, cyber security, the military, finance, law and technology. That is not office-page decoration; it explains the product's shape. Investigators want provenance and room for judgment, engineers want repeatability, and compliance buyers want the answer packaged for an audit. The company's public culture stresses integrity, collaboration and continuous learning, while its India team pages show cricket leagues, outdoor trips, charity events and Diwali celebrations. The serious craft here is learning how another discipline treats evidence - then making that practice legible inside software.

The boundaryPublic intelligence is not omniscience

The approach works only when the risk leaves a discoverable trace and the investigator has enough seed information to identify the right subject. A private agreement, an unreported offence or records sealed behind an inaccessible jurisdiction will not materialise because an algorithm searched harder. Sparse digital footprints can mean low risk, careful privacy or simply poor coverage. They are not proof of virtue.

Employment screening also carries a special hazard: irrelevant personal material can create bias rather than insight. Neotas says its research is ethical and transparent, but the customer still needs a lawful purpose, proportionate scope and a decision policy that distinguishes job-relevant risk from somebody's lawful private life. In other words, the method is strongest when the question is precise, the consequence merits the intrusion and a human can explain the result.

That sober limit is why the company's hybrid position makes sense. A pure database is fast but looks backward. A traditional investigation is rich but expensive. A generative answer is fluent but may be impossible to defend. Neotas is trying to occupy the useful middle: fast until the facts become strange, deep when the stakes demand it, and documented throughout. The product is not the red flag. It is the path from suspicion to a decision somebody is prepared to sign.