Watch · 1 minute
AIDA Labs in one minute
Captioned, no sound needed.
Consulting from the people who build the tools
More than 15 years of program delivery inside banks and other regulated businesses, now paired with AI that builds fast inside firm limits. We plan, design and deliver data, analytics, Salesforce and portal projects, and we build working software on free, open-source tools, hosted in Canada, at a fraction of the usual cost.
Every box opens the details. Every sample link opens something that is running today.
Three ways to work with us
Consulting services
How we use AI, and where we don't
What we can put to work for you
Our own products, running today on fictional data. Each one can be used out of the box, tailored or rebuilt as part of an engagement. See all our products ›
Open the live demo ↗
How an engagement runs
Proudly Canadian, Practically Priced
Start with a conversation
Guardrails we already use
These run today in AIDA's own systems, and every one is a pattern we bring to client work.
Our founder
Mike MacLeod
Mike has spent more than 15 years leading technology programs as a project manager, business analyst and solution architect, most of it inside banks and other regulated businesses. At Tier1 Financial Solutions (2018 to 2024) he delivered financial-services CRM programs and spent two years delivering Anti-Money Laundering software and analytics on the Alessa platform, running up to five compliance projects at once. Before that he was a solution architect at MASS Engines, and most recently he led delivery as a Senior Project Manager at HSS Business Solutions.
His work includes a $3.1 million, 18-month CRM rollout for Scotia Global Banking and Markets, and taking over Barclays' flagship engagement and moving it from red to green within weeks, with reporting and an issue-management system he built himself.
In 2026 he built AIDA on his own, from the ground up, using AI: a job-search and recruiting platform with its own database, scoring engine, workflow automation and secure client portal, running on free and open-source software for about $15 a month. He wrote the guardrails its AI works under, and they are the basis of how AIDA builds for clients.
He then built the AIDA Labs product suite: project portals with installable apps, a live AML Intelligence Hub with five dashboards and 50 KPIs, Executive Database Analytics, SaaS Project Management with Project Plans and Blueprints, and the data model for Project Boards.
A proud Canadian, Mike designed the whole architecture to keep data in Canada. It is hosted on a secure server in Quebec under Law 25, with backups and syncs at a second site in Toronto, and runs on free and open-source software rather than US-owned platforms.
Credentials
- Project Management Professional (PMP), 2026
- Salesforce Certified Administrator, 2012
- Salesforce Certified Platform Developer I, 2012
- B.Sc., Dalhousie University
Start with one of our products as it stands: the Intelligence Hub, SaaS Project Management, an AIDA-base with Executive Database Analytics, or a client portal.
We load your data, set your names, users and roles, and adjust the configuration that matters. You pay for a short set-up, not a build.
Best for: teams that need a proven tool quickly and are happy to work the way it already works.
We adapt a product to how you already work: your KPI definitions and thresholds, your terminology, your release and sign-off process, your branding, and the reports your leaders already read.
It can run as a service on our Quebec server, or be installed on your own servers or cloud.
Best for: organizations with an established process that want our tools to fit it, not the other way round.
A solution designed and built from the ground up: the data model, data collection APIs, the database, the dashboards or portal on top, and the documents that go with them.
- Purely mechanical: rules, a well-designed database and deterministic processing. Many problems are best solved this way, and it costs nothing to run per transaction.
- AI-enhanced: AI used only where it adds real value, with every call designed to use as few tokens as possible, so running costs stay low and predictable, and a person stays in control of every change.
You own the result: the code, the database design and the documents.
We find out which numbers matter, where they come from and how to make them trustworthy, then build what delivers them:
- an inventory of reports, owners and sources
- pipelines with quality gates and daily reconciliation
- a reporting mart (facts and dimensions) designed to move to Azure, Databricks or Fabric
- dashboards where every number opens the records behind it
Proven on our AML Intelligence Hub: five live dashboards and 50 KPIs on a PostgreSQL database. See the sample ↗
A KPI dictionary written in plain language: the definition, the formula, the source, the owner and the threshold, agreed in a workshop with the people who will be held to it.
Each KPI is built as a versioned SQL function, tested against today's reports in a parallel run, and certified by its owner before it is published. After that there is one trusted number, and anyone can open the records behind it.
We help you decide where AI belongs, and where rules and a good database do the job better and cheaper. Then we plan and deliver it with three parts working together:
- People decide, review and run anything that touches production
- Mechanical rules and scoring do the repeatable work the same way every time
- AI reads, drafts and explains, sparingly, inside limits a person set
We design the guardrails in from day one: AI never runs commands against production, deleting is off by default, dashboards are read-only, and every AI use is approved, logged and kept within the data boundary.
Salesforce Certified Administrator and Platform Developer I, with financial-services CRM programs delivered for banks, including a $3.1 million, 18-month CRM rollout.
We configure, extend and report on Salesforce, and we build data that moves cleanly between Salesforce and our free-software products, so you can use each where it fits best, or move off per-seat licences where you don't need them.
Client portals, public websites and installable apps (they work offline and install on any phone, tablet or computer), built on free software and hosted on our server in Quebec.
Examples running today: the AI-Enhanced Job Seeker Portal ↗, the AI-Enhanced Recruiter Portal ↗ and the sample project portal ↗.
Project Management Professional (PMP) delivery for data, analytics and CRM programs: the charter, the plan, releases and sprints, the RAID log, steering committees and sign-offs.
Track record: taking over a flagship banking engagement and moving it from red to green within weeks, and running up to five compliance projects at once.
We run projects on our own SaaS Project Management: the plan and the Project Blueprint on one page, readable by everyone, with no per-seat licences. See a sample plan ↗
Our AI-enhanced portals are designed to use three kinds of intelligence strategically, so each one strengthens the others and covers their weaknesses:
- Mechanical does the heavy, repeatable work: collecting, cleaning and scoring every job or candidate with the same versioned rules, every time. It never gets tired and never guesses.
- AI is used sparingly, where reading and explaining language helps: summarizing, drafting and explaining a score. It never decides on its own.
- People review the results before a client sees them, and make the final call.
Why no AI or agent can reach our architecture: AI writes code and commands, but a person reads and runs every one. AI sessions have no access to the databases, servers, containers or schedulers, deleting is off by default, and the databases sit behind an outbound-only tunnel and a service-token gate the AI never sees. Human intervention is the barrier, by design.
No AI runs inside it. Every number comes from plain, versioned SQL functions on your data, so results are deterministic, repeatable and auditable. We use AI only to accelerate its design, build and testing, which lets one experienced person deliver what usually takes a team, and keeps your cost down.
That applies to the Intelligence Hub, Database Analytics and Insights, Project Blueprints, Project Boards and every custom database we build. AI helps us design, write, test and document them. Once built, they run on PostgreSQL, PocketBase, nginx and Linux, with no model in the loop.
We start by laying the whole project out: proposal, charter, scope, schedule, releases, risks, decisions and sign-offs, each a deliverable you can open, and a plan with every task, milestone and the critical path.
You see the controls before the build starts, and so does everyone else who needs to sign off.
AI is how one experienced person delivers what usually takes a team. It writes, tests and documents quickly, but it works under strict rules:
- a person reads and runs every command that touches production
- AI sessions can't delete files without a person approving it
- dashboards and portals use read-only access
- no AI or agent can reach the database, servers or architecture on its own
Results are checked against the source and against today's reports in a parallel run. Variances are explained, and each KPI is certified by its owner before it is published.
Run it as a service on our secure server in Quebec, with backups in Toronto, or take it in-house.
Everything uses free software and open formats (PostgreSQL, CSV and JSON), with a runbook and data dictionary, so it moves to Azure, Databricks, Salesforce or your own servers whenever you choose.
Our production server is in OVHcloud's data centre in Beauharnois, Quebec, run by its Canadian company, Hébergement OVH Inc. Backups and data syncs go to our second site in Toronto.
The company that runs our Quebec server is Canadian and part of a French group, not a US provider, so the US CLOUD Act does not reach the data we store. Today, visits pass through Cloudflare for protection and speed; it does not store your data, and we are moving to a direct connection so traffic stays in Canada too.
Law 25 sets strict rules for personal information: a person responsible for privacy, privacy impact assessments, incident reporting and clear consent.
We design for it from the start: only the data the job needs, access by role, read-only dashboards, retention and classification rules per data zone, and every change made by a person and logged.
Platforms such as Salesforce charge for every user, every month, so the bill grows with your team. We build on PostgreSQL, PocketBase, nginx and Linux instead.
You pay for the solution and for hosting it, not for each person who uses it.
AI accelerates design, coding, testing and documentation, so one experienced person can deliver what used to take a team. The guardrails stay the same, and the savings are passed on to you.
We look at one thing with you: a report people don't trust, a project that's slipping, a manual process, or an AI idea.
You leave with a plain-language view of what's wrong, what it would take to fix, and whether you can do it yourselves. Sometimes the right answer is a better spreadsheet, and we'll say so.
A small, fixed-price piece of work with a clear deliverable:
- a Project Blueprint and plan for your project
- a KPI dictionary for your top ten numbers, agreed with their owners
- a prototype dashboard or AIDA-base on a copy of your data
If it's useful, it becomes the first release. If not, you keep it.