Skip to content

THE COMPANY BEHIND OPENADAPT

About OpenAdapt.

OpenAdapt is an AI automation company that builds and maintains workflows across existing systems for operations teams.

We combine hands-on implementation with software your team can inspect and reuse. Start with one workflow, agree what a good result looks like, and build from there.

Discuss your workflow

FOUNDER

Richard Abrich

Richard is an AI architect and the founder of OpenAdapt. He helps businesses put AI automation into production and develops the software behind OpenAdapt. He holds a Master of Applied Science from the University of Toronto.

MLDSAI Inc. builds OpenAdapt in Toronto, Canada. The project’s public code dates to 2023. Richard leads the work, with open-source contributors improving the software through public issues and pull requests.

Meet the project through its work

CUSTOMER WORKFLOW

A physician’s monthly audit. A practical reason to automate.

Dr. Victor Abrich’s monthly RVU audit requires evidence from clinical records and report spreadsheets. OpenAdapt collects the evidence from Cerner PowerChart and reconciles the findings into a workbook for his review.

The physician checks the findings before the billing follow-up. That division of work guides our implementation process: automate the repeated steps and give the operator a result they can check.

Read the workflow and evidence

The case study’s financial figures remain preliminary, pending customer confirmation.

WHAT OPENADAPT DOES

From a repeated task to a working implementation.

01

Workflow assessment

We map the process with your operators, measure the current workload, and agree the result checks. You get a defined scope before committing to a build.

Start with the workflow
02

Build and integration

We use APIs, conventional code, document processing, and AI where each fits. OpenAdapt’s computer-use software handles steps that require the application interface.

See the delivery process
03

Deployment and maintenance

We test normal cases and exceptions with your team, then train the people who operate the workflow. Continued support covers the monitoring, changes, and exceptions agreed in your scope.

Plan ongoing support
04

Reusable automation software

OpenAdapt compiles demonstrated GUI workflows into deterministic, locally executable programs. The platform connects recording, execution, and result evidence for teams that build and operate their own workflows.

Explore the platform

A defined result before the build

We agree the baseline, acceptance criteria, and operator responsibilities for one workflow. Those checks determine whether the implementation is ready for daily use.

Checks beyond the last click

The Flow runtime can check the intended effect against an independent result source. When the configured checks can’t establish the result, the workflow stops for review.

An engine you can inspect

The compiler, runtime, and local launcher are MIT-licensed. Your engineers can inspect the code and run it locally; Cloud and implementation services are commercial offerings.

A deployment that fits your data

We agree the execution environment and data boundary with your technical owner. Sensitive observations stay inside that boundary, with review before approved material leaves it.

WHO WE BUILD FOR

Operations teams with work that repeats across systems.

Start with a workflow that has a named owner, enough volume to justify automation, and a result you can verify.

  • Healthcare administrationEvidence collection, document review, and report reconciliation.
  • BPO and service teamsRepeated tasks across client portals and desktop applications.
  • Document-heavy operationsIntake, field checks, and approved updates to business systems.
  • Software and engineering teamsWorkflow execution inside an existing product or agent integration.

The first step is a 30-minute workflow discussion. Tell us which systems you use, how often the task runs, and how you check the result today.

Before a paid build, we agree the scope, price, acceptance criteria, delivery schedule, and people responsible. Your proposal sets the communication channel and response expectations. We plan operator training before handover and scope ongoing maintenance separately.

For software issues, use GitHub. Community discussion takes place on Discord.

COMPANY REFERENCE

Key facts

Name
OpenAdapt
Business
AI automation software and implementation services
Started
2023
Founder
Richard Abrich
Based in
Toronto, Ontario, Canada
Core offering
Build, verify, deploy, and maintain repeated business workflows
Software
MIT-licensed local engine; commercial Cloud. Platform directory
Service pricing
Quoted for an agreed workflow scope; maintenance agreed separately
Software pricing
Current plans and terms
Contract and support
Scope, schedule, communication, and response expectations set in the engagement agreement
Company profiles

Do we buy software or an implementation?

You can use the open-source engine yourself or work with us on a paid deployment. The Services page explains delivery; the Platform page explains the software.

Do our systems need an API?

An API can be the best integration path when one is available. For steps that require a browser or desktop interface, we assess OpenAdapt’s computer-use capabilities against your workflow.

Can sensitive work stay in our environment?

We agree the data boundary and execution environment before deployment. Native and remote desktop work runs in the customer-controlled environment specified for that workflow.

How much does an implementation cost?

We quote after reviewing the workflow, its systems, and its acceptance criteria. The build and ongoing maintenance have agreed scopes; software plans are listed on the pricing page.

What should we bring to the first call?

A description of one repeated task, the systems involved, and an estimate of the time it takes. Explain what can go wrong and how someone checks that the work is complete.

START WITH THE WORK

Show us the task your team keeps repeating.

Discuss your workflow