Canadian AI Research Lab

Autonomous agents for sports analytics

We build AI systems that discover features, run experiments, and evaluate models without human intervention. Turning raw sports data into predictive intelligence.

563
ML Features
531
Experiments
13
Repositories
2,054
Commits
745+
R&D Hours
Platform

End-to-end autonomous sports analytics

Agentic AI replaces the manual hypothesis-experiment-analysis cycle. Our agents discover features, design experiments, and evaluate models across multiple prediction targets.

01

Autonomous Feature Engineering

AI agents generate predictive features from multi-source sports data autonomously. A novel dual lag schema captures both weekly and per-game temporal patterns, scaling feature spaces far beyond what manual engineering achieves.

02

Automated ML Experimentation

End-to-end pipeline that generates hypotheses, designs experiments, tunes hyperparameters via Bayesian optimization, and evaluates results across multiple prediction targets and player positions — no human in the loop.

03

Multi-Source Data Integration

A medallion architecture ingests, validates, and joins heterogeneous data sources with automated entity resolution achieving high cross-source matching accuracy.

Architecture

Medallion architecture

Modular medallion architecture. Each layer independently upgradable, every boundary enforced by automated validation.

LayerRepoPurpose
BronzethebronzedaleRaw data ingestion with Pydantic validation
SilverthesilverdaleCross-source joins, entity resolution, data quality
GoldthegolddaleML-ready feature engineering
DiamondthediamonddaleML training, tuning, inference, backtesting
QuantumthequantumdaleAutomated feature synthesis
DashboardthedashboarddaleInteractive web UI and analytics dashboard
AI Infraclaude-systemAgent workflows, ACE methodology, memory systems
AI Infraopencode-systemSwarm delegation, ML model auditing
EvolutionaryinfinitedaleAutonomous AI code evolution framework
Research & Development

Active research areas

We work on genuinely unsolved problems at the intersection of agentic AI and sports analytics. Failed approaches are part of the process — that's what real research looks like.

  • Uncertainty 01
    Autonomous AI Feature Discovery
    Can AI agents autonomously discover generalizable predictive features from heterogeneous multi-source data? We're exploring the ceiling of automated feature generation and cross-season generalization.
  • Uncertainty 02
    AI-Driven Hypothesis & Experimentation
    Can AI agents reliably replace the human hypothesis-experiment-analysis cycle? We're building systems that generate, execute, and evaluate experiments end-to-end across player positions and prediction targets.
  • Uncertainty 03
    Multi-Source Data Integration at Scale
    Can heterogeneous sports data be automatically unified with ML-grade fidelity? Automated validation and entity resolution across structurally different sources.
  • Uncertainty 04
    End-to-End Agentic ML Framework
    Can the full ML lifecycle — ingestion through evaluation — be orchestrated by AI agents via a single CLI-driven pipeline? Multi-task, multi-target coordination remains an open problem.
  • Uncertainty 05
    Novel Prediction Targets
    Can ML models predict non-standard targets — boom/bust probability, ceiling distributions, ownership patterns — using quantile regression and multi-task learning?
  • Uncertainty 06
    Evolutionary AI Code Optimization
    Can AI systems evolve their own code through competitive selection and mutation? We're building a framework where agents compete and improve through multi-perspective judged evaluation.
Target Market

Built for the sports industry

The sports analytics market is projected at $12B by 2028, growing 25% annually. We serve organizations that need data science capability without the overhead of a dedicated team.

Teams

Professional Teams & Front Offices

Player evaluation, draft scouting, salary cap management, and in-game strategy for NHL, CFL, MLS, and other North American leagues. Most teams below the top tier lack dedicated data science staff.

Agencies

Player Agencies & Talent Management

Data-driven player valuations for contract negotiations, career development planning, and identifying undervalued talent across leagues and positions.

Analytics

Sports Media & Analytics Firms

Predictive analytics for broadcasts, content platforms, and data companies. Companies like Sportradar and Stats Perform seek better models and feature pipelines to enrich their products.

Team

Two technical founders

Both under 35. Complementary expertise in ML, data engineering, and AI infrastructure.

Olivier Maguire

Co-Founder — ML & Data Engineering

Lead architect of the medallion pipeline and ML experiment framework. Designed the feature engineering schema and pioneered the agentic AI methodology for autonomous experimentation.

British Columbia, Canada

Garry Airiants

Co-Founder — AI Infrastructure & Evolutionary AI

Built the AI development infrastructure — agent workflows, swarm delegation, ML auditing, observability. Designed the evolutionary AI framework where agents compete, mutate, and evolve code.

Alberta, Canada
Accessibility

Accessibility-first, not accessibility-later

Our platform is designed to be the first sports analytics tool with comprehensive accessibility from the ground up. Personal motivation — a founder's close friend has MS.

WCAG 2.1 AA

Full compliance across all interactive features — data tables, charts, controls, dashboards.

Motor Impairment

Full keyboard navigation, voice control, large click targets, switch access compatibility.

Visual Accommodations

High contrast mode, screen reader optimization, scalable interface elements.

Cognitive Load

Progressive disclosure, simplified views, clear information hierarchy.

Canadian Innovation

Supported by Canadian R&D programs

Bootstrapped, under 35, and leveraging the Canadian innovation ecosystem to accelerate deep R&D while maintaining full ownership.

Wage Subsidy

NRC IRAP

Industrial Research Assistance Program support for continued R&D in agentic AI and autonomous ML experimentation.

Tax Credits

SR&ED

Scientific Research & Experimental Development tax credit program. Documented technological uncertainties and systematic investigation as supporting evidence.

Startup Financing

Futurpreneur

Both founders under 35. Eligible for Futurpreneur startup loan and BDC matching. Mentorship alongside capital.

Contact

Get in touch

Interested in our platform, our research, or exploring a partnership?

No Strings Labs Ltd.

BC Corporation, est. 2023

info@nostringslabs.com
olivier@nostringslabs.com
garry@nostringslabs.com
British Columbia & Alberta, Canada