Skip to content

Resolve

Resolve is the autonomous investigation agent of SenseOn. It is one of the specialist agents that Horus, the AI orchestrator of SenseOn, coordinates. Resolve analyses the security cases that the SenseOn platform raises. It collects evidence from multiple data sources and gives clear verdicts that evidence supports. Thus, your team uses less time on repetitive triage and more time on important work.


What does Resolve do?

Each time SenseOn raises a high or critical case that the SenseOn SOC manages, Resolve automatically takes the case and investigates it. It does this in the same way as an experienced SOC analyst. Resolve collects context, examines historical patterns, and queries threat intelligence. Then it makes a structured verdict with a confidence score.

The result is a clear recommendation for each case:

  • Escalate. This case must have your attention. Resolve flags proven threats, suspicious activity, and all activity that it cannot dismiss as benign.
  • Close as False Positive. The detection fired incorrectly. The observed activity does not match what the rule was designed to detect.
  • Close as Benign True Positive. The detection fired correctly, but the activity is really harmless. Examples are a sysadmin who runs legitimate maintenance scripts, scheduled backup software, or routine IT asset discovery.
  • Needs More Info. There is not sufficient data for a confident decision. Resolve tells you exactly what data is absent, so that your team can follow up.

Each verdict includes the evidence for it, a confidence score, and recommended next steps. Thus, your analysts get all the information that is necessary to do a check of the finding or to act on it.


How does it work?

Resolve uses a structured investigation process for each case:

  1. Evidence collection. Resolve collects data from multiple sources. These sources include case details, detection rule context, host and network observations, historical case patterns, and threat intelligence feeds.
  2. Analysis. Resolve analyses all the evidence together to get a complete picture of the activity. It considers the purpose of the detection rule, and whether the observed behaviour matches known benign patterns. It also considers the past outcomes of similar cases.
  3. Verdict delivery. Resolve makes a structured verdict with a confidence score, key evidence, and recommended analyst actions.

What data sources does Resolve use?

Resolve uses a large set of context for each investigation:

  • Case details. The full case timeline, observations, and metadata from Hunt Lab from SenseOn.
  • Detection rule context. What the rule that fired is designed to detect, and why it fired.
  • Historical patterns. The past outcomes of similar cases, for your organisation and across the wider SenseOn customer base.
  • Host and network telemetry. Related observations from the same hosts, users, and IP addresses.
  • Threat intelligence. Enrichment from multiple threat intelligence sources. These sources cover IP reputation, domain analysis, file hashes, and known indicators of compromise.

Confidence scoring

Each verdict has a confidence level. Thus, your team knows exactly how much trust to give the recommendation:

Confidence What it means
High Strong historical evidence supports this verdict. Clear detection context with no contradictory signals. You can act on this with high assurance.
Medium Moderate evidence for the verdict. Some ambiguity, or a single signal without corroboration. A quick review is useful before you act.
Low Limited historical data, or contradictory signals. Use it as the start of more investigation.

How does Resolve handle security testing?

Resolve is cautious by design. If it detects activity similar to penetration testing, red team exercises, phishing simulations, or vulnerability scanning, it will always escalate. This is also true if the activity seems to be expected or scheduled.

Only your team can say if security testing is authorised. Resolve will never assume that test activity is benign. Thus, you always have visibility into these events.


What are the benefits?

Faster case resolution. Resolve automatically investigates the cases that analysts reviewed manually before. This decreases the time to a verdict from hours to minutes.

Consistent, evidence-backed decisions. Each verdict uses the same structured investigation process, the same data sources, and the same standards. Thus, there is no variation between analysts or shifts.

More time for high-value work. Resolve handles the high-volume, repetitive cases. This lets your security team focus on genuine threats, proactive threat hunts, and strategic security improvements.

Full transparency. Each verdict includes the evidence that Resolve collected, the reasons for the decision, and clear next steps. Your team always has all the information, and it makes the final decision.

Continuous learning from history. Resolve considers the past outcomes of similar cases across your organisation and the wider SenseOn customer base. Thus, its pattern recognition improves with each case.


Frequently asked questions

Does Resolve replace the SenseOn SOC? No. Resolve handles the repetitive, high-volume cases. Thus, the SOC can focus on complex investigations and threat hunts.

Can Resolve take actions in my environment? Resolve is an analysis and recommendation engine. It investigates cases and gives verdicts. It does not make changes to your environment, block users, or isolate hosts.

How accurate is it? Resolve has a strict quality standard, with more than 99% agreement with the decisions of experienced analysts. Each verdict includes a confidence score to help you prioritise.

What if Resolve gets it wrong? The design of Resolve includes safety. If Resolve is not sure, it escalates the case and does not close it. High-confidence auto-close thresholds are under joint management and strict oversight. Any unexpected result triggers an immediate review.

Does it work with my existing SenseOn setup? Yes. Resolve works with your current SenseOn deployment. No additional configuration is necessary. It reads cases from your environment and gives verdicts together with your current workflows.