Skip to main content

RPA and process automation, with the ROI calculated before we start

UiPath, Power Automate and Python robots. We quantify the gain in hours before building, and we turn down automations whose maintenance cost exceeds the gain.

Model
Fixed price or embedded
Duration
3 to 8 weeks per process
Expected outcomes
Hours given back to teams who then do something other than re-keying
Expected outcomes
A documented, maintainable robot estate rather than hidden debt

Who it is for

  • Finance and HR departments with massive manual re-keying
  • IT departments facing automation demand without capacity to serve it
  • Insurance, banking and mutual back-offices with manual cross-application flows
  • Companies whose ERP does not talk to the rest of the IT estate

Who it is not for

  • Processes run fewer than twice a month
  • Processes whose rules change every month
  • Automations that would bypass an internal control

Deliverables

  • Timed map of the existing process with real volumes
  • ROI calculation: hours saved per year against build and maintenance cost
  • Robot built, versioned in your repository, with usable logging
  • Documented exception handling: what the robot does when things go wrong
  • Execution dashboard (success rate, average duration, exceptions)
  • Manual fallback procedure if the robot is unavailable
  • Training for business users and level 1 support

How it runs

  1. 01

    Qualification and timing

    We watch the real process, run by the person who runs it, and time it. Self-reported timings are wrong by a factor of two in both directions.

  2. 02

    ROI decision

    Annual gain against three-year total cost including maintenance. If payback exceeds eighteen months we recommend not automating, in writing.

  3. 03

    Build and hardening

    The robot is written to fail cleanly: error recovery, exception queue, named alerting. A robot that fails silently costs more than the manual process.

  4. 04

    Go-live and follow-up

    Four weeks running in parallel with the manual process, then cutover. Real gain measured at three months against the original promise.

Stack

  • UiPath
  • Power Automate
  • Power Platform
  • Python
  • Playwright
  • n8n
  • Camunda
  • REST / SOAP
  • SAP GUI Scripting

How the assignment is structured

Three ways to engage this service. The figure is set at the end of framing, once the scope is written: it is the only way to commit firmly without pricing in a risk you would end up paying for.

Automation assessment

fixed price, up to 5 processes

You have a candidate list but no prioritisation

  • Mapping and timing of 5 processes
  • ROI calculation per process
  • Sequencing recommendation
  • Firm quote for the selected robots

Fixed-price robot

fixed price per process

One stable, high-volume process

  • Robot build and hardening
  • Exception handling and logging
  • Execution dashboard
  • Business and level 1 support training
  • 90-day warranty

Robot estate maintenance

subscription, 6-month term

An existing estate that breaks with every application update

  • Execution monitoring and alerting
  • Selector breakage fixes
  • Adaptation to version upgrades
  • Monthly availability report
Get your scope priced

Frequently asked questions

Hasn't RPA been superseded by APIs?

Where an API exists we use it: more robust, cheaper to maintain. RPA remains justified in front of applications that expose nothing, vendor-locked packages, or timelines where opening an API would take eighteen months of committee work. We will tell you which case you are in.

Do you really turn projects down?

Yes. Around a third of the processes submitted to us fail the ROI threshold, usually because they are too infrequent or too unstable. We document that in the assessment, which is yours to keep even if we build nothing.

What happens to the robot if the target application changes?

It breaks. That is the reality of RPA. Which is why we ship resilient selectors, immediate breakage alerting, and why the maintenance contract exists. An automation sold without a maintenance budget is an automation sold in bad faith.

Other services

Twenty minutes is enough to know whether we are useful

No sales deck. You describe the need, we say whether it is in scope, at what price and on what timeline. If it is not for us, we say so during the call.