Why investigation context should live where the data does, and what's new in Agilence Analytics 2.15.
TL;DR: Agilence Analytics 2.15 introduces Comments, a new capability that lets investigators anchor notes directly to POS transactions — making investigation context queryable, reportable, and durable. The release also includes Configurable Date/Time Dimensions, giving analysts the flexibility to summarize measures by alternate date fields. Both are live now and demoing at NRF PROTECT 2026 in Grapevine, Texas, June 8–10. Visit us at Booth #1209, or request a demo.
Three months after a flagged transaction, an analyst opens it again and asks the question every LP team has heard a hundred times: Why is this marked as Not Fraud?
Sometimes the answer comes back quickly. More often, the analyst who tagged it has rotated shifts, moved to another company, or simply doesn't remember. The reasoning lives somewhere outside the system — in an old email, a Slack thread, a spreadsheet only one person updates, or in someone's head. By the time the same pattern resurfaces, the context is gone, and the investigation effectively starts over.
That's the knowledge problem.
Investigation context almost never lives with the transaction it belongs to. Walk through any loss prevention team's real workflow and you'll find the reasoning behind decisions scattered across email threads, shared Excel trackers, group chats, and the institutional memory of whichever analyst happened to handle the case. None of those places are queryable. None of them are tied to the data. And none of them survive a turnover event.
Retail trade turnover hovers around 60% annually, according to the U.S. Bureau of Labor Statistics. LP and AP teams are not immune. Investigators rotate, analysts move on or forget prior cases, and the deep context they've built up, such as the reasoning behind a tag, the relationship with a particular store, the history of a recurring pattern, leaves with them. What looked like a clear tagging convention or a deliberate exception becomes, six months later, a mystery for the next person to inherit.
Three things consistently break down when investigation context lives outside the data.
Repeat investigations. When a similar pattern resurfaces, the team often reinvestigates from scratch — not because the original work was poor, but because the original reasoning was never captured in a place anyone could find. The same store, the same associate number, the same transaction signature, worked twice. Time that should compound is spent recreating it.
Audit defensibility. When legal, HR, or compliance asks why a transaction was tagged the way it was (or why an investigation was closed without escalation), screenshots of Slack threads and forwarded emails are not a defense. Teams need a documented record that ties the decision to the data it was made about: who made it, when, and why. That kind of trail is hard to assemble after the fact and easy to maintain if it's built into the workflow from the start.
Knowledge that walks out the door. When an experienced investigator leaves, every undocumented reason behind every tagged transaction leaves with them. The team replaces a person but cannot replace the context. New hires inherit dashboards full of decisions they cannot explain, and the team spends weeks or months rebuilding pattern recognition that already existed.
The fix isn't another tool to manage. It's a category change in how teams think about the work itself: investigation context is a first-class data asset, not a byproduct. Notes, reasoning, and decisions should live next to the transactions they're about, not in adjacent systems where they decay.
Done that way, every investigation a team runs today makes the next one faster. The knowledge compounds. New investigators inherit institutional memory instead of orphaned dashboards. Audit questions resolve quickly because the answers are right where the data is. And the team's collective expertise becomes part of the platform, not a function of who's currently on staff.
This is the thinking behind Comments, a new capability in the 2.15 release of Agilence Analytics. Investigators can add notes directly to POS transactions — the reasoning behind a tag, what was learned in a call to the store, why a case was or wasn't escalated. Comments are anchored to the transaction they describe, permission-controlled, and preserved with a full audit trail.
Where it really earns its keep is in reporting. Comments are queryable through new fields including Comment Count and Comment Exists, which means teams can do things like surface the transactions carrying the most comments, filter a refund report to only transactions an investigator has weighed in on, or measure how much investigative activity is happening on a particular store, region, or category over time. Notes stop being sticky notes and start behaving like data.
Comments in 2.15 is deliberately scoped. It supports POS transactions today, with plain-text notes. It's human-authored context, captured where it belongs, creating a foundation that everything from richer collaboration features to AI-assisted analysis can be built on later. Building the data layer first matters: a comment thread an LP team writes today will still be useful, and still be searchable, when future capabilities arrive on top of it.
The 2.15 release also introduces Configurable Date/Time Dimensions, giving analysts the flexibility to summarize measures by alternate date fields. Late-night transactions can roll up to Business Date instead of calendar date; cash discrepancies can be analyzed by Posted Date versus Actual Date; RFID data can compare Last Seen verses Lost Date to identify operational issues with inventory movement especially around high-shrink categories; and case data can be summarized by Incident Date, Created Date, or Closed Date depending on the question being asked. With Configurable Date/Time Dimensions, previously hard to find fraud is easily discovered. A small change, but with a real impact on how cleanly reports reflect operational reality.
Agilence is at NRF PROTECT 2026 in Grapevine, Texas, June 8–10. Stop by Booth #1209 to see Comments in action and talk through what investigation context could look like in your team's workflow.
Can't make it? Request a demo, and we'll walk you through it.