What's New in Nexus: Agentic Investigations, 47 Skills and Evidence-Linked Findings

Describe your question naturally and develop the investigation through follow-up requests. In the new Nexus, a case conversation is where you state an objective — the subject, the research question and any relevant timeframe or context — rather than a search box for one query at a time. Nexus interprets the request, works out what it needs to establish first and carries the investigation through as one connected workflow: source discovery, evidence retrieval, relationship analysis and reporting.

There are three ways to start. Pick a Skill from the Skill Library when you have a specific research objective; describe a broader investigation in your own words when you do not; or run a direct lookup when a single record is all you need. Whichever you choose, the investigator stays in control. You review supported findings, possible matches, contradictions and information gaps, then direct the next step.

Follow-up questions build on what has already been found. Ask Nexus to focus on one candidate, compare two conflicting accounts or extend the research into a related subject, and the new material joins the same case alongside the evidence that prompted it.

Nexus previously offered a set of individual lookups — web search, social discovery, WHOIS, DNS, IP, breach, offshore and payment identifiers — each answering one narrow question and leaving the investigator to connect the results. The new Skill Library replaces that one-lookup model with 47 investigative Skills. A Skill is a focused workflow tailored to a specific research objective: it states what question it answers, what you receive and what it cannot establish, and it returns cited findings rather than raw records.

The library is organized into 7 categories: Research and Analysis; Identity and Public Presence; Business, Financial and Risk Research; Media, Publications and Public Activity; Internet Infrastructure and Exposure; Web3 and Cryptocurrency; and Relationship Analysis. Together they bring public identity discovery, business and sanctions research, multimedia analysis, digital exposure and Web3 intelligence into one investigation workspace.

Much of the coverage is new to Nexus: company directories and official registries, sanctions and PEP screening, image matching, video transcripts, academic and patent footprints, app and advertising research, public forum mentions and cryptocurrency wallet intelligence. Existing strengths such as social discovery across 200+ platforms remain, now delivered as Skills with the same evidence discipline.

The October 2026 update adds seven Skills to the Business, Financial and Risk Research category. EU Sanctions Measures reviews European Union sanctions records with their legal instruments, restrictions and effective dates. PEP Offices and Associates reviews politically exposed person records, the offices held and the associates the records attribute. Sanctions Ownership Exposure traces the ownership and control relationships that sources attribute to risk-screening candidates, and Vessel Sanctions Screening checks a vessel by IMO number, MMSI or name.

Regulatory Enforcement Review covers concluded enforcement and regulatory actions in US futures and derivatives industry records, and Wanted-Person Record Check covers the wanted-person records of Hong Kong’s anti-corruption authority. Risk Change Review compares the two most recent saved risk assessments for a subject and explains what changed. Each new Skill keeps identity similarity and record relevance apart, states its coverage limits and is included on both Pro and Agency. The same update gave Cross-Entity Correlation its own Relationship Analysis category.

Evidence links open retained records with available source details and context. Follow an original public link where provided, save useful material to Findings, or add supported evidence to a selected Agency graph.

This is what separates an agentic OSINT platform from an AI summary. A summary gives you an answer; Nexus gives you the answer and the material behind it, so you can check the claim yourself before it enters a report. Findings are presented as supported findings, possible matches, contradictions or information gaps, and an unsuccessful search does not establish that information is absent.

Answers also render tables and structured text, so timelines, candidate comparisons and identifier lists are readable in the case conversation and carry cleanly into briefs and reports.

On Agency, add supported evidence records to a selected graph and explore connections between people, organizations and digital identifiers. You can also create nodes and connections manually. Review what supports each relationship; adding a connection to a graph does not verify it.

Nexus treats relationships with the same discipline as findings. A shared username, a similar image or a common hosting arrangement is recorded as an association to examine, not as proof of identity, ownership or control. Cross-Entity Correlation, an Agency Skill, researches and assesses relationships across people, organizations and digital identifiers, distinguishing direct evidence from inference.

Review concise assessments or export selected case material into a structured PDF report. BLUF Situation Brief presents the bottom line first, followed by the evidence, assessment and limitations that support it, and is one of the three core Skills included on every plan. Deep Public Research and News and Adverse-Media Timeline produce cited reports and sourced timelines that separate established events, allegations and conflicting accounts.

Research outputs distinguish supported findings, assessments, contradictions and gaps. Pro and Agency can export selected case records to structured PDF reports, including available custody entries where enabled and a SHA-256 hash of the prepared case data. Review the content and completeness before use; neither the hash nor the report establishes legal admissibility.

Keep investigation conversations, findings, notes and relationship graphs organized around a case. Cases, notes, findings, tasks and reports were already the backbone of Nexus. In this release, every investigation output — conversation answers, Skill results, evidence records and graph additions — is connected to the case it belongs to, so a follow-up question weeks later starts from the same evidence rather than from scratch.

Investigation notes preserve your analysis and observations alongside the case, with end-to-end encrypted notes on the Agency plan. Progress is shown consistently while a Skill runs, and evidence navigation takes you from a citation to the evidence record to the original source and back to the case without losing your place.

Since 1 October 2026, new accounts choose between two plans. Pro costs €59 a month and includes 43 of the 47 investigative Skills and 150 credits a month; Agency costs €89 a month and includes all 47, adding Cross-Entity Correlation, Breach Pattern Analysis, IP Abuse History and IP Threat Signals, with 300 credits a month. Unused credits carry over without a cap, and prepaid credit packs remain available when you need more.

Before the update, Pro included 100 credits a month and Agency 200. Teams, a multi-seat plan for investigation teams, is coming soon. Each Skill page on this site lists the plans that include it.

Pro and Agency continue to support custody records when enabled at case creation and structured PDF exports of selected records. Review the available trail before relying on it; a case-data hash does not verify the PDF or its conclusions. Social discovery across 200+ platforms, encrypted Agency notes and the case workspace remain available.

Nexus is still built and operated by Next Sight d.o.o. in Ljubljana, Slovenia, an EU company, and it remains a tool for lawful, ethical open-source research. Nexus supports investigation and assessment, not automatic declarations of identity, ownership or wrongdoing.