Founder-led, in Richmond Hill, Ontario.
MAS-AI maps how a business runs and builds the software, automation or AI that closes the gap. The person who diagnoses the problem is the person who builds the system.
The company
- Legal name
- MAS-AI Technologies Inc.
- Founded
- January 2026
- Based in
- Richmond Hill, Ontario, Canada
- Founder
- Masoud Masoori, Founder and CEO
Who does the work
- Founder
- Masoud Masoori scopes and builds every engagement. He builds AI agent, retrieval and automation systems; his earlier career was in IT support, quality assurance and technician roles.
- Study
- Artificial intelligence, in the Artificial Intelligence graduate certificate program at Seneca College, 2025.
- How
- The founder works with AI coding and review assistants and reviews their work. Projects are built with automated tests and checks.
- Public work
- Daena, a governed multi-agent AI platform (source-available under the Business Source License 1.1), and MCP Switchboard and MergeLoop (open source, Apache 2.0).
How we work
The person who maps your business is the person who designs and builds the system. No handoff to a team you have not met.
You own it
The code, the data and the accounts are yours, as set out in the contract.
Scope first
A fixed scope and price before any build starts.
People approve
AI never takes an irreversible action without a person.
Handover
Documentation and a walkthrough your team can run with.
Checks before release
Tests and deterministic checks run before anything ships. A change is done when a check says so, not when it looks right.
AI output is evaluated
Where a model produces an answer, we measure it against defined cases before it reaches your people or your customers.
Audit logs
Automated actions leave a record of what was done, when, and on whose authority.
Human approval
A person approves any action that cannot be undone.
Secrets in a vault
Keys and passwords live in a vault. They are never written into code or shared documents.
Documented handover
Every build ends with written documentation and a walkthrough, so your team can run it without us.
Patent application
Neural-Backed Memory Fabric (NBMF)
No. 64/020,421, filed March 2026
A provisional application secures a filing date. It is not a granted patent, and we do not describe it as one.
Programs
- Google for Startups Cloud ProgramCloud credits
- Microsoft for StartupsAzure credits