SPUR treats education as infrastructure. A country that wants to build with AI needs a pipeline of people who understand it - from curious kids, to working professionals, to founders shipping real products. We run that pipeline as a public good: free courses, open compute access, and a studio path for the learners who want to keep going.
“Judge a person by their questions rather than by their answers.”
Paraphrased from Voltaire
Three principles shape everything we put in front of a learner.
Our first course is written so a ten-year-old can follow it and a skeptical parent can finish it in an afternoon. Jargon is optional. No signup paywall, no credit card, no course fee.
Every course ends with something you built. Not a certificate-for-showing-up - a working project. Text generator, document Q&A, a small agent. Whatever you finish, you keep.
Our course platform runs on SPUR's Canadian compute. Learners who graduate into serious work get sovereign GPU access through the venture studio. No US cloud lock-in.
Start where you are. Each course is free, interactive, and self-paced.
A gentle, kid-first curriculum covering what AI is, where it shows up in everyday life, and how to talk to it safely. Short lessons featuring Chip the AI helper, with fun activities to do together. No tech setup required.
A gentle intro to AI. What it really is (a very practiced word-predictor), how it learns, what it can and can't do today, and how to talk to it. Short lessons with short quizzes. If you've felt intimidated by all the AI talk lately, start here.
The full intro course covering AI basics, prompting, building with AI (RAG, agents, APIs), making it yours (fine-tuning, LoRA, evals), and shipping to production. Ends with a capstone project you submit for SPUR review. Pass it and earn a verifiable certificate plus optional compute credits on SPUR GPUs.
Real prompting patterns that hold up, retrieval-augmented generation end to end, embeddings and vector search, long-context strategies, multi-turn conversation state, AI for data and writing workflows, multimodal basics, and shipping your first AI feature.
Agents that take action, fine-tuning and LoRA, evaluation that catches regressions, structured outputs and schema validation, multi-agent orchestration, retrieval at scale, cost and latency engineering, prompt security, caching and batching, and shipping production AI pipelines.
Transformers from the inside, post-training (RLHF, DPO, synthetic data), pretraining (data, compute, architecture), Mixture-of-Experts, efficient attention and long context, inference optimization, interpretability, the agentic frontier, and AI safety and governance.
How to actually use AI in a real job. Picking a daily-driver model, the prompts that separate pros from tourists, building work agents (no-code and code paths), research and decision-making with AI, communication, automation, integrations with Slack/Notion/Gmail, safety and privacy at work, and team rollout.
Pick problems AI actually helps with, design UX for uncertainty, write evals that matter, price AI features, and ship real products. Distinct from the engineering tracks - this is the product-shipping angle.
The honest practitioner guide to creative AI tools. Claude and GPT for writers, Midjourney/Flux/SDXL for images, Runway/Kling/Sora for video, Suno and ElevenLabs for audio. Real workflows, real licensing pitfalls.
Ollama and llama.cpp for daily driving · vLLM/SGLang/TGI in production · quantization that works · hybrid patterns with frontier fallback · compliance · the real cost math vs Anthropic/OpenAI. Aligned with SPUR's Canadian-sovereign compute.
Slurm, MPI, InfiniBand · CUDA programming · profiling with nsys/ncu · FSDP, DeepSpeed, Megatron · cost accounting. Runs real workloads on SPUR's own clusters for the capstone.
Qubits, superposition, entanglement · H/X/CNOT gates · Grover and Shor · VQE/QAOA for NISQ devices · Qiskit hands-on · post-quantum crypto and where quantum ML is actually real in 2026.
Ideation and validation · customer interviews that teach you something · pre-seed/seed fundraising mechanics (SAFE, SR&ED, IRAP) · co-founders and first hires · first ten customers and GTM · connects to the SPUR venture studio path.
Create a free teacher account, spin up a classroom for any of the courses, and share a 6-character invite code. Your students sign up, enter the code, and their progress appears on your dashboard as they work. Built for K-12 teachers, universities, bootcamps, and corporate L&D teams. Need something tailored? SPUR will work with you to build a custom curriculum for your program - domain, tools, pacing, capstones, all of it.
See all courses → | Stats refreshed live from learn.spuric.com
For learners whose circumstances shouldn't decide whether they can study AI.
SPUR awards scholarships each year to students pursuing AI, HPC, or quantum research who face financial barriers. Awards cover tuition, compute credits, and (for selected recipients) on-site internships at the SPUR Innovation Centre.
Learn about scholarshipsThe full AI Foundations course is free globally. Learners in regions where local compute is expensive or unavailable can apply for SPUR GPU credits to complete their capstone - awarded on merit of their project proposal, not nationality.
Request compute for my capstoneFor graduates who finish the course and want to turn their capstone into a company.
The top capstone projects each quarter are invited into the SPUR Venture Studio - an intensive program that pairs you with SPUR engineers, gives you sovereign compute, and helps you go from "working demo" to "company people pay for." Equity-light, founder-friendly terms.
How the Studio worksNot every great capstone needs to become a company. Learners who want to keep building but don't want the startup path can apply for ongoing SPUR compute allocation for research, open-source projects, or community tools. Projects that benefit the Canadian AI ecosystem get priority.
Apply for compute