Content Observability for Kofax + Tungsten Automation

Your Kofax estate is going somewhere. Make sure it lands.

Stay on Capture and KTM, move up to TotalAgility, or leave the ecosystem altogether — all four are defensible. None of them are safe to attempt without knowing what the current system actually does. Reveille measures it, agentlessly, before anything changes.

Get Started See the four options
Tungsten Automation Technology Partner Agentless · zero footprint On-premises · cloud · hybrid
TODAY Your estate Kofax Capture · KTM TotalAgility · supporting servers Stay on Capture and KTM Run it with less vendor dependency Move up to TotalAgility Measure both sides of the move Already on TotalAgility Assure what it replaced Another IDP platform Baseline before you go ONE MEASUREMENT LAYER, ALL FOUR PATHS Throughput · queue depth · batch aging · field accuracy · service levels
At a glance · Kofax + Tungsten Automation

Migration assurance for the Kofax install base — the facts.

  • Architecture
    Agentless, zero footprint — nothing installed on Kofax or Tungsten Automation servers
  • Coverage
    Over 100 pre-built application-aware tests, over 180 dashboard metrics
  • Products
    Capture, Transformation Modules, TotalAgility, Communication Server, Front Office Server, VirtualReScan, Import Connector, Reporting
  • Partnership
    Tungsten Automation Technology Partner
  • Baseline captures
    Throughput, queue depth, batch aging, batch exceptions by batch class, field accuracy, module processing speed, licence and page use
  • Service levels
    Monitor availability, service level and system availability reporting against target and minimum thresholds
  • Integrations
    Splunk, ServiceNow, Datadog, PagerDuty, Azure Monitor, OpenTelemetry-compliant tooling
  • Outcomes
    95% Service Level Assurance attainment · 50%+ less downtime · 20+ hrs reclaimed weekly
Why migrations go sideways

The platform migrated fine. Nobody can prove the work did.

Most Kofax and Tungsten Automation migrations are not scoped badly. They are scoped without evidence — because the one thing nobody wrote down is what the old environment actually did. Six months later there is an argument about whether the new system is slower or less accurate, and nothing to settle it with. Platform SLA is not workflow SLA, and during a transition that gap is at its widest.

01

Nobody wrote down what “working” meant

Teams know the environment runs. Very few know it in numbers — documents per hour on a Monday morning, how long a batch really sits, what field accuracy holds at across a quarter. Those figures are easy to capture while the system is live and effectively impossible to reconstruct once it is gone.

02

The party doing the migration grades the migration

The vendor recommending the move supplies the dashboard that reports on it. The integrator delivering it writes the completion report. Both are fine at their jobs and neither is a disinterested witness. Your vendor cannot grade its own homework.

03

The old numbers disappear at cutover

The moment the legacy environment is decommissioned, the comparison set goes with it. Whatever was not measured beforehand becomes a matter of opinion — usually litigated between IT and a line of business that is absorbing the rework.

Your options

Four ways this goes. We are useful in all four.

Reveille has no preference about where a Kofax environment ends up. The measurement is the product, and it earns its keep whichever path a customer takes — including the one where nothing moves at all.

01 / PathNot funded, not ready, not this year

Stay on Capture and KTM

The renewal is the decision point, but the business case to move isn’t there yet — or the team that knows the batch classes is the same team that would have to rebuild them.

What Reveille doesRun the estate with less dependence on a support queue that has gone quiet: catch failures earlier, clear routine batch errors without a ticket — clear a batch task error, release a batch hold, retrigger a batch step, cycle a stalled service — and keep the service levels the business already expects. Proactive actions run in automatic, conditional, or audit-only mode, so nothing moves in production until you say it can.
02 / PathCommitted to the move, or scoping it now

Move up to TotalAgility

Validation forms, validation rules, foldering and document-review logic get re-created on the destination side. The question that usually goes unasked is how anyone proves the result matches what was there before.

What Reveille doesCapture the baseline while the legacy environment is still running, then measure throughput, queue behaviour and service levels on both sides through the parallel run.
03 / PathMigrated already, living with the result

Already on TotalAgility

The move is done. What is missing is an independent view of whether the new environment performs the way the old one did — and a way to see Core Worker throughput, job processing and export service behaviour without asking the platform to grade itself.

What Reveille doesOver 100 application-aware tests and over 180 dashboard metrics across TotalAgility and the supporting servers, feeding the incident tooling already in place.
04 / PathEvaluating the wider IDP market

Leave the ecosystem

A renewal is a natural moment to ask whether this platform is still the right one. Every vendor in the evaluation will show you their own numbers; none of them will show you yours.

What Reveille doesBaseline the environment you have today so the evaluation runs on evidence. Where the destination is a platform Reveille covers, both sides get measured the same way.
The system snapshot

What a picture of your environment actually contains.

Over 100 pre-built, out-of-the-box tests exercise the critical points of the Tungsten Automation platforms, feeding over 180 dashboard metrics tracked over time. Wizards build a working monitor from a library that already understands the architecture — which is what makes it realistic to stand a baseline up on a production estate mid-decision.

01 / Tungsten TotalAgility9 test groups

Tungsten TotalAgility

Connectivity, platform, database, import connector, clients, licensing, WCF service metrics, reporting and performance — including Core Worker tasks, threads, locked activities, Search and Matching Server and Export Service.

02 / Kofax Capture8 test groups

Kofax Capture

Queue counts, overall batch count, batch time in system, batches in error by batch class, attended and unattended module speed, remote site currency and batch class download, licensing volume, station use, license expiration and backup license server, email and web services import connectors.

03 / Transformation Modules (KTM)5 test groups

Transformation Modules (KTM)

Queue counts, field accuracy — too high on rejection and incorrect, too low on correct — module processing speed, statistics database, and the full Real-Time Transformation Interface.

04 / VirtualReScan3 test groups

VirtualReScan

Scanner make, model and firmware, pages-scanned rate, sheets between feeder, multi-feed and jam errors, camera health cadence, roller health, uniformity and clarity.

05 / Communication Server8 test groups

Communication Server

Message queue depth, processing speed, queue page counts, license registrations, application and messaging server health, channels, mail system size and archive entries.

06 / Front Office Server6 test groups

Front Office Server

In-process transaction count and age, release error count, server and application transaction rates, MFP job submission, user inbox transaction count and queue time.

07 / Front Office Server MFP2 test groups

Front Office Server MFP

Device reachability, device status and device errors, plus device-specific checks per MFP model in the estate.

08 / Import Connector6 test groups

Import Connector

Messages waiting by media type, failing counts, license state, connector status and storage, email inbox and folder item counts measured against expected thresholds, import folder file age and total count, import archive folder access, SMTP server, fax FOIP port and disk space.

09 / Reporting4 test groups

Reporting

Central site and application sender services, ETL agent, message queue performance, database size, and inbound and outbound active and processed message counts.

See the full Tungsten Automation coverage
Before, during, after

One measurement, carried across the cutover

A baseline is only worth something if the same thing is measured on the other side. Reveille records the legacy estate while it is still the environment your team knows, then keeps measuring through the parallel run and past cutover — so a dip is attributable instead of arguable.

100%98% 96%94% MEASURED BASELINE RANGE MINIMUM COMMITMENT CUTOVER Below minimum for 11 days Attributable, because the before was measured. LEGACY ESTATE · MEASURED BEFORE ANYTHING CHANGES DESTINATION PLATFORM
Legacy estate, measured Destination platform Cutover Illustrative — shape of a real reporting period, not a customer result
Minimum commitments

Turn the baseline into something you can hold somebody to

This is where a measurement stops being a report and starts being leverage. Reveille’s service level analysis reports every monitored resource against a target and a minimum — the same structure a migration statement of work needs, and the same structure an auditor will ask for.

Service level analysis · score card structure
MeasureTargetMinimumActual
Service levelset from baselinecontractual floormeasured
System availabilityset from baselinecontractual floormeasured
Average resolution timeminutesminutesmeasured
Allowed down timeminutesminutesmeasured
Monitor availability
(minutes monitor available ÷ time period monitor running) × 100
Service level
(minutes available ÷ (minutes monitored − allowed down time)) × 100
System availability
(minutes available ÷ time period) × 100

Each test carries its own reporting flag, so the service level reflects the transactions the business actually cares about rather than every ping in the estate. Tests can be grouped so that high-availability configurations get credit when one node covers for another. And each test can carry a cost — revenue lost to downtime, cost of repair — which is what turns queue depth and batch aging into a number a CFO recognises.

Service Level Assurance

You can’t negotiate a service level you’ve never measured.

Whichever direction your Kofax estate moves, the commitment you end up with is only as good as the evidence behind it. Reveille is the independent record — before, during, and after.

Customer outcomes

What Content Observability customers typically see

Typical results across the Reveille customer base. Individual outcomes vary with estate size, platform mix, and how much of the environment is brought under measurement.

95%
Service Level Assurance attainment across ECM, IDP, and automation platforms
50%+
typical decrease in downtime and ticket volume
20+
hours per week reclaimed from firefighting
1,000+
purpose-built tests across every supported ECM, IDP, and automation platform
Where we measure both sides — and where we don’t

Reveille covers ABBYY, Box, Hyland, IBM, Microsoft, OpenText, Tungsten Automation and UiPath. If your destination is on that list, both sides of the move get measured the same way, by the same layer, from one console.

If it isn’t — and there are good platforms that aren’t — Reveille cannot measure the far side. The baseline of the environment you have today is still yours, and it is still the thing a new vendor or integrator should be held to. We would rather say that here than on the first call.

Part of a broader platform

Tungsten Automation is one of many platforms Reveille observes

One observability layer across every major Enterprise Content Management (ECM), Intelligent Document Processing (IDP), and automation platform — which is what makes measuring a cross-vendor migration possible at all. It also means the measurement does not stop at the capture boundary: captured content is released into a repository, and Reveille checks those output targets are online and redrives repository connections when they are not. Capture through to commit, in one view.

Tungsten TotalAgility · Capture · KTM Hyland OnBase · Alfresco OpenText Documentum · Capture IBM FileNet · Datacap · CMOD ABBYY FlexiCapture · Vantage UiPath RPA assurance Microsoft SharePoint · M365 Box Content cloud
Questions

Kofax and Tungsten migration assurance, answered

What is Reveille migration assurance for Kofax and Tungsten Automation?
Reveille migration assurance is agentless Content Observability applied to a platform transition. Reveille monitors, alerts, self-heals, and reports across Kofax and Tungsten Automation environments, establishing a measured baseline of what the current system does — throughput, queue depth, batch aging, extraction accuracy, service levels and user activity — before anything changes. That baseline becomes the acceptance criteria for the migration and the yardstick the destination is measured against.
Does Reveille require agents on Kofax or Tungsten Automation servers?
No. Reveille is a zero footprint management application: it does not change the state of, or maintain a persistent connection to, the target system. Nothing is installed on Kofax or Tungsten Automation servers, which means a baseline can be stood up on a production estate mid-migration without adding platform risk.
Which Kofax and Tungsten Automation products does Reveille support?
Reveille supports Tungsten Automation Capture, Transformation Modules, TotalAgility, Communication Server, Front Office Server, Front Office Server MFP, VirtualReScan, Import Connector and Reporting. Tungsten Automation RPA is covered by a separate Reveille technical overview. These are the products formerly branded Kofax Capture, KTM, KTA and Kofax RPA.
What can Reveille measure in a Kofax Capture or KTM environment?
In Capture, Reveille measures queue counts, overall batch count, batch time in system, attended and unattended module processing speed, remote site currency and batch download, licensing volume and station use, and email and web services import connectors. In Transformation Modules, Reveille measures queue counts, field accuracy — flagging when accuracy is too high on rejection and incorrect results or too low on correct results — attended and unattended module processing speed, and the full Real-Time Transformation Interface including request counts, failure counts and average request duration.
Does Reveille work with Tungsten TotalAgility deployed in the cloud?
Yes. Reveille operates against Tungsten Automation platforms running on-premises or in the cloud, and Reveille itself runs on AWS EC2, Google Compute Platform and Microsoft Azure Windows instances as well as on-premises. Because the approach is agentless, the measurement works the same way against a hosted TotalAgility environment as against an on-premises one.
Can Reveille measure extraction accuracy in Transformation Modules?
Yes. The Transformation Modules wizard includes a field accuracy test that flags when accuracy is too high on rejection and incorrect results, or too low on correct results, alongside queue counts and module processing speed. This is the measurement most often missing from a migration business case, because it is the one that determines how much manual rework the business absorbs after cutover.
How quickly can Reveille be deployed before a migration?
Wizard-driven templates build a working Tungsten Automation monitor from a prebuilt library that already understands the platform architecture — there is no agent installation and no custom integration work. In practice a baseline can be running well inside the window between a migration being scoped and a cutover being scheduled.
What is a migration baseline, and why does it matter?
A migration baseline is a measured record of what the current environment actually does, captured while it is still running: volumes, throughput, queue depth, batch aging, extraction accuracy, service level attainment and user activity. It matters because it is straightforward to capture beforehand and effectively impossible to reconstruct afterwards. Without it, questions about whether the new environment is slower or less accurate cannot be settled with evidence — only with opinion.
Can Reveille measure both the old and the new environment during a parallel run?
Yes, where both sides are platforms Reveille covers. Migrations of this size commonly run the legacy and destination environments in parallel for an extended period, and Reveille observes both from a single console, so throughput and service levels on either side are measured the same way by the same independent layer.
What if we are migrating to a platform Reveille does not cover?
The baseline of your current environment is still yours. Reveille covers ABBYY, Box, Hyland, IBM, Microsoft, OpenText, Tungsten Automation and UiPath. If your destination is outside that list, Reveille cannot measure the far side of the move — but it can still measure the near side, and the resulting record of volumes, accuracy, throughput and service levels is exactly what you need to hold a new vendor or systems integrator to a commitment.
Is Reveille a Tungsten Automation partner?
Yes. Reveille is a Tungsten Automation Technology Partner. Reveille's role is not to argue for or against a migration — it is to make the outcome measurable whichever direction a customer chooses.
How do I get started?
Start by requesting a demo. A Reveille engineer will walk through your current Kofax or Tungsten Automation environment, identify which measurements matter most given where you are in the decision, and outline what a baseline would capture. Because the architecture is agentless and wizard-driven, most environments can be under measurement quickly and without change control on the target platform.
How does Reveille compare to SolarWinds, Splunk, Datadog, or Nagios during a migration?
SolarWinds, Splunk, Datadog and Nagios are general-purpose infrastructure monitoring tools — they watch servers, networks, operating system metrics and log volumes. They do not understand Kofax or Tungsten Automation as applications, so they cannot tell you whether a Capture batch aged past its service level, whether Transformation field accuracy shifted, or whether Core Worker task throughput changed after a cutover. Reveille produces that content-layer signal and feeds it into the tools you already run — including Splunk, ServiceNow, Datadog, PagerDuty, Azure Monitor and OpenTelemetry-compliant tooling — rather than replacing them.
What does Reveille measure that the migration team cannot?
Independence is the structural difference. The vendor recommending a migration, and the integrator delivering it, both supply the reporting that says how it went. Reveille is an observability layer that is not built, sold or operated by the platform it measures, so the record of before-and-after service levels is authored by something with no stake in the answer. Platform SLA is not workflow SLA, and during a transition that gap is at its widest.
Do we have to rebuild our monitoring when we migrate?
No. Reveille monitors the same categories of issue on every platform it covers — batch health, aging, exceptions, licensing and channel processing — using tests purpose-built for each platform underneath. The tests differ because the mechanics differ; the operational discipline, the thresholds you tuned and the service level reporting carry across. A migration becomes a platform change rather than a monitoring restart, which is also why standing monitoring up once ahead of the move costs less than standing it up twice.
Can Reveille act on a problem, or does it only report it?
Reveille includes a remediation component with over 70 proactive actions it can take in response to a detected error — clearing a batch task error, releasing a batch hold, retriggering a batch step, setting batch priority, cycling a Windows service or process, running a PowerShell recovery script, or redriving a repository connection. Each action runs in automatic, conditional or audit-only mode, so a team can see what Reveille would have done before letting it do anything. During a migration that matters twice over: fewer tickets on the legacy estate, and a record of every intervention on both sides of the cutover.
Next step

The content layer is where your business runs. Reveille makes sure it holds.

Start with a picture of the environment you have today — what it processes, how fast, how accurately, and where it strains. Whatever you decide next, you will decide it on evidence.

Get Started → Explore Reveille for Tungsten Automation