LABSISUPER INTELLIGENCE

INDEPENDENT AI LAB · ALBERTA, CANADA

Local roots.
Useful AI.

A founder-led lab in Sherwood Park, exploring local language models, agent systems and mathematical reasoning for Canadian businesses.

Explore the lab See what we’re building
An imagined architectural AI lab with a server rack, robot workbench, coding monitor, mathematical board and research notebook.NEURAL LABBROWSER EXAMPLE / 01A simple rule.
Something to explore.F(n) = F(n−1) + F(n−2)0   1   1   2   3   5   8   13   21Open the sequence explorer

Choose a research station.

LAB VISUALISATION / 01

An imagined space for real work in progress.

Try the browser experiments

LAB SUPER INTELLIGENCE

Canadian AI.
Built with curiosity.

The research and development side of Opcelerate Neural Inc., led by Andrés Garcia Quirarte in Sherwood Park, Alberta.

I’m working on my own models, agent systems and ways to build with AI. The Lab brings that work into one place: questions to explore, systems to develop and things we can learn together.

Our direction is practical: more useful local AI, clearer control over tools and data, and experiments that can be explained. Model capability and hardware support will be documented through specific tests.

Meet the person behind the Lab

ON THE WORKBENCH

Five ways
to move the work forward.

These are active areas of exploration. Each track starts with a question and a next step.

01 / THE MODEL BENCHResearch in progress

Local models

Language models closer to your work.

Exploring model development, adaptation and local inference for Canadian businesses. The aim is useful AI with more choice over where it runs and what it can access.

Inside this research track

The question

How much useful capability can we bring to practical, affordable hardware?

The next step

Record the model, hardware, task quality, memory use and speed together. Publish comparisons only when the test can be explained and repeated.

Explore the station
02 / THE SYSTEMS BENCHDevelopment in progress

Agent factory

Give every agent a job.

Building ways to connect specialised agents, tools and a human reviewer. The work includes task routing, reusable workflows and deciding what an agent is allowed to do.

Inside this research track

The question

Can a small group of well-scoped agents make a real workflow easier to run?

The next step

Test clear handoffs, visible records and a way to stop or correct the work. Start with a narrow task before expanding access.

Explore the station
03 / THE BUILD BENCHDevelopment in progress

Coding systems

From an idea to working software.

Exploring coding assistants and the software around them: interfaces, tool connections, testing and review. A useful system needs more than generated code.

Inside this research track

The question

What helps a person check, improve and maintain an AI-assisted build?

The next step

Review changes, run relevant checks and make the outcome understandable to the person using it. Keep release decisions with a person.

Explore the station
04 / THE REASONING BOARDExploratory research

Mathematics

Patterns worth questioning.

Exploring mathematical sequences, structured reasoning and ways to evaluate predictions. Small, understandable experiments help separate an interesting pattern from a reliable method.

Inside this research track

The question

Can we describe the rule, test it on unseen cases and understand when it fails?

The next step

Start with explicit assumptions and simple baselines. The sequence explorer below is an educational browser example, not a model benchmark.

Explore the station
05 / THE RESEARCH NOTEBOOKOngoing learning

Research & learning

Make the work easier to understand.

Connecting experiments with practical learning for owners and teams. Research notes can explain the question, the method, what happened and what still needs work.

Inside this research track

The question

What does someone need to know before they try this in their own business?

The next step

Prepare clear notes, demonstrations and project-specific guidance. Public model releases and benchmark reports will be identified when available.

Explore the station

A SMALL WINDOW INTO THE THINKING

Try an
experiment.

Simple, working examples you can explore here. These run in your browser and illustrate ideas, rather than the performance of an AI model.

BROWSER WORKBENCHLOCAL INTERACTION · NO SIGN-UP

MATHEMATICS / INTERACTIVE EXAMPLE

A rule.
A sequence.
A question.

Choose a sequence and change how many terms you see. The rule is explicit, so you can inspect every value.

F(n) = F(n−1) + F(n−2)

Bar heights show relative values within this sequence.

0 · 1 · 1 · 2 · 3 · 5 · 8 · 13 · 21 · 34 · 55 · 89

Deterministic mathematics. This example does not use an LLM.

THE RESEARCH APPROACH

Make it useful.
Show how it works.

Our ambition is capable, practical AI for Canadian businesses. Clear evidence is how we get from an idea to a claim.

01

Start with a real task.

Choose something that matters to a person or a team. Define a useful outcome before choosing the model or tool.

02

Test on named hardware.

Record the conditions and the trade-offs. “Local AI” needs a specific setup, rather than a promise about every computer.

03

Keep the boundary clear.

Label concepts, browser examples, prototypes and released work. Publish performance comparisons with a method and limitations.

Background reading: llama.cpp and local inference ↗

THE PUBLIC NOTEBOOK

Room to build.
Room to learn.

This page is the starting point for the Lab’s public work. Future project notes can include screenshots, demonstrations, videos, methods and results.

LAB NOTE / 001

Opening the Lab.

Local models. Agent workflows. Coding systems. Mathematical questions. Five research tracks, and a public place to explain what we’re exploring.

Current stage: research and development in progress. No model weights or performance benchmark reports are published on this page yet.

Our Canadian AI perspective Read how we approach useful AI for Canadian businesses. Lab results will be labelled separately.

A FEW USEFUL ANSWERS

About
the Lab.

Curious about a specific project?
Text Andrés ↗

What is Lab Super Intelligence?

It is the founder-led research and development side of Opcelerate Neural Inc., based in Sherwood Park, Alberta. The work focuses on local AI models, agent systems, coding, mathematical reasoning and practical learning.

Is this a Canadian AI research lab?

The Lab is an independent initiative within an Alberta company, led by Andrés Garcia Quirarte in Sherwood Park. It is a place for applied experiments and development, with Canadian business needs in mind.

Will your model run on any computer?

Hardware support is something to test for each model and task. Memory, model size, software and the required speed all matter. The goal is broader access to useful local AI; compatibility will be documented for specific tested configurations.

Can I download a model or see benchmark results?

This page introduces work in progress. It does not currently offer model weights or published benchmark results. The interactive examples are small browser demonstrations, clearly separated from model research.

Does local AI mean all my data stays on my computer?

That depends on the complete setup. Models can run locally while other tools still connect to external services. Data access, storage, network connections and any cloud dependencies need to be agreed and checked for each project.

Can my company work with the Lab?

Yes, you can discuss a use case, research collaboration or a scoped pilot with Andrés. We first agree what you want to learn, what data and hardware are available, and how a useful result would be assessed.

Can we meet in person?

Andrés is based in Sherwood Park. The first 30-minute meeting is free, in person by appointment, by phone or on video.

SHERWOOD PARK · ALBERTA · CANADA

What would you
like to test?

Bring a problem, a research question or an idea for local AI. Let’s work out a useful first experiment.

Talk with Andrés
Call Andrés Text Andrés

Free 30 minutes · In person, phone or video

IMAGINED LAB / RESEARCH IN PROGRESS

The question

The next step

Read the research track Try a browser experiment

LAB SUPER INTELLIGENCE · MEET ANDRÉS

Let’s talk
about your idea.

Arrange a free 30-minute conversation about local AI, a workflow or a research pilot for your business.

Call Andrés(778) 488-8113Text Andrés(825) 459-3324
Send my project details

In person by appointment in Sherwood Park, by phone or on video.

Prefer email? Write to Andrés ↗