Claims Management is live. Automate freight claims from detection to dispute resolution.

Claims Management is live.

Explore it here

What Shipped at Arvist: August 2026

Summary: Arvist shipped major improvements to warehouse quality control in August 2026, achieving 99% packing accuracy on live orders, adding real-time error detection for packers, and automating safety reporting through AI. The company also introduced proof-of-load documentation via mobile capture and video recording to resolve shipping disputes with timestamped evidence, shifting focus from preventing false positives to catching the costly false negatives that humans miss during busy warehouse operations.

ON THIS PAGE

This month was about tightening the loop between detection and action: packing stations got more accurate and more visible in real time, safety reporting got automated, and load documentation got a second (and soon a third) way to happen automatically.

Let’s get into it.

Pack station accuracy reaches 99% 📦

What we shipped

Packing station inspection is now hitting up to 99% product identification accuracy, and it’s not a lab number. It’s coming from live client installations, processing real orders on real warehouse floors every day.

  • Live accuracy, not a benchmark: the 99% figure is measured on real orders at active client sites, not a controlled test environment
  • Fast time to value: sites are reaching this accuracy level within days of initial installation
  • Self-learning Product Catalog: a new catalog that improves its own matching as it sees more products move through the station
  • Improved inspection handling UI: a cleaner interface for packers to work from as inspections happen (more on this below)

Why it matters

An accuracy number only means something if it holds up outside a demo. Reaching 99% product identification on live orders, within days of go-live, is the bar Arvist is setting for AI-powered warehouse QC as packing station inspection rolls out to more sites: fewer missed errors, faster resolution, more confidence in every shipment.

Impact

  • Up to 99% product identification accuracy on live packing station orders
  • Accuracy reached within days of installation, not weeks
  • A self-learning catalog that keeps improving with use

Catch errors before the box ships ✅

What we shipped

The inspection app now gives packers real-time visibility into every product as it’s counted.

  • Live per-product counters: expected vs. actual quantities track on screen as the shipment comes together
  • Instant mismatch flags: overages, shortages, or wrong products are surfaced the moment they happen
  • Clear expected-vs-detected pop-ups: packers see exactly what was expected and what was actually detected, no guesswork

Why it matters

An error caught while the box is still open costs seconds to fix. The same error caught after the shipment leaves costs a customer complaint, an investigation, and a claim. Moving that detection to the moment of packing, instead of after the fact, is what turns a QC system from a record-keeper into something that actually prevents mistakes.

Impact

  • Fewer errors reach the dock
  • The errors that do happen get caught and fixed in seconds, not discovered days later in a customer complaint
  • Packers get a clear, on-screen answer instead of relying on memory or a paper pick sheet

AI-generated safety reports, on demand 🤖

What we shipped

Say goodbye to manually compiling safety data. The new Safety AI reporting tool builds comprehensive, stakeholder-ready reports in seconds.

  • Set the scope: facility, zone, or camera
  • Set the audience: the agent tailors the report accordingly
  • The agent does the rest: querying incident data, analyzing trends, and assembling a polished executive summary

Why it matters

Safety reporting has traditionally meant someone pulling incident data by hand, building the same charts every reporting cycle, and writing up trends from scratch. The new reporting tool reads like it was written by a safety analyst: clear metrics, trend context, and plain-language takeaways, generated automatically from the live camera network. Less time building reports, more time acting on what they reveal.

Impact

  • Stakeholder-ready safety reports generated in seconds instead of compiled by hand
  • Reports scoped to exactly the facility, zone, or camera that matters for the audience
  • Executive-summary quality output, generated automatically from live camera data

Proof of load, captured in the moment 📸

What we shipped

Damage or shortage disputes after a shipment leaves are notoriously hard to resolve: was it loaded correctly, or damaged in transit? Arvist now gives you documented proof either way, two ways:

  • On-demand mobile capture: the forklift driver snaps load images directly from the Arvist inspection app on a mobile device, as pallets are placed
  • End-to-end video capture: the full loading process records automatically from start to finish, no manual step required

Forklift-mounted camera capture is next, so load documentation happens automatically as part of the normal loading motion, with zero extra steps for the team.

Why it matters

A dispute that used to come down to one party’s word against another’s now comes down to a timestamped record. Whether the capture is a quick mobile snap or a full video of the load going on, there’s a verifiable record tied to the shipment, so when a claim comes in, there’s evidence to resolve it fast instead of a guessing game.

Impact

  • A verifiable, timestamped record tied to every shipment, captured at the moment of loading
  • Two capture methods live today, mobile and full video, with forklift-mounted capture in development
  • Disputes resolved on evidence instead of on who argues harder

Worth a watch: the metric that actually matters 🎯

Arvist CTO Bazyl Horsey sat down with Joe Leucht of Hardesty Technologies this month to talk about what really moves the needle in warehouse QC.

False positives get the attention: they’re annoying, visible, easy to point at. The harder, more valuable problem is false negatives, the damage or wrong barcode a human on a long shift would likely miss. Those quiet misses are what turn into costly escapes. The bigger idea: instead of adding headcount to manually pull evidence during peak season, a system that’s always watching the gaps humans struggle to catch can do the job of multiple people, and prove it, with a KPI that translates the gap closed into verifiable, dollar-denominated ROI. That’s the shift from “cool tech” to hard numbers on the P&L.

🔗 Watch the full conversation on LinkedIn

Looking ahead ⏩

Packing station inspection keeps rolling out to more sites, and the self-learning Product Catalog will keep sharpening as it sees more volume. On the safety side, this is just the first version of AI-generated reporting; expect more report types and more ways to customize scope and audience in the months ahead.

If you want to see any of this in person, visit the Arvist Experience Center or talk to our team.

Related content

What Shipped at Arvist: July 2026 This month was about closing the loop on claims and making the platform easier to configure and act on in […]

...
4 mins

Every warehouse team knows the feeling. You catch the error, you know you’re right, and then you spend the next three weeks proving it to someone […]

...
4 mins

Please wait...