VP of Machine Learning @Shopify | Building Agents | Bringing AI to entrepreneurs | Bush Pilot | Maple Syrup Producer https://xitter.cf/t.co/haSezvLkIm
Belwood, ON
Joined January 2010
- Tweets926
- Following35
- Followers4.3K
- Likes42.4K
Going to chat about productionizing agents, post training, and inference optimizations on @nebiusai live at #NVIDIAGTC Berlin in 35 minutes or so. Tune in!
Nebius Live at NVIDIA GTC Berlin | Day 2 xitter.cf/i/broadcasts/1RKZzBdvl…
Talking at 3pm at @nvidia GTC Berlin on how @Shopify post-trains open source models and optimizes inference using Nvidia's models and tools. Read our blog post about it here: shopify.engineering/open-mod…
Nice work by @cmazzaanthony and @Shopify ML team on our flywheel. Cody is walking through our continual learning framework at @PyTorch conf that optimizes our task specific agents beyond frontier models at lightning speeds and at a fraction of the cost.
Now: @Shopify’s continual learning loop with PyTorch and @vLLM_project takes the stage at #PyTorchCon 2026 with Cody Mazza-Anthony @cmazzaanthony.
Using Buyer Profile as an example, Cody walks through how a calibrated judge flags low-scoring outputs in production. Critics diagnose each failure, and a replay with a critic hint is scored again by the judge. Accepted repairs expand and improve the training data.
Cody shows how a small specialist model can outperform a frontier model on this task, then turns to optimizing inference through prompt compression.
Watch the keynote livestream: events.linuxfoundation.org/p…
Andrew McNamara retweeted
Loved working with you @Drewch - thanks for the support
Had the pleasure of embarking on the wild journey of Maluuba with @spisallyouneed building agents over 15 years ago before people were even talking about Siri or ML. Pumped about Sky 001, the team is cooking!
Had the pleasure of embarking on the wild journey of Maluuba with @spisallyouneed building agents over 15 years ago before people were even talking about Siri or ML. Pumped about Sky 001, the team is cooking!
Introducing SKY-001, the first Engineering World Model.
On CAD and BIM tasks, it beats GPT-6 Astra and every frontier model we tested:
-> 51-125x faster
- > 150-210x cheaper
-> Higher IOU than all frontier models
SKY-001 lets you dream a physical object before you build it.
18 months of work. Built by the team out of Maluuba's deep learning lab (acq. by @MSRResearch).
The first ‘AI General Engineer’ will be the backbone for Autonomous Manufacturing.
Andrew McNamara retweeted
Shopify does train large LLM - frontier-class - already. Just not in text space - in Action space. Shopify trains the largest HSTUs in the world!
Andrew McNamara retweeted
Turn any existing Agent into the world's best personal shopper for you.
The new shop skill: shop.md/
Add the skill & connect your Shop account & finally find the thing that you always wanted
Andrew McNamara retweeted
3rd best employer on the planet.
Proud of every single person who makes Shopify Shopify 💚
Introducing the World's Best Employers.
To determine the list, Forbes and Statista surveyed more than 300,000 employees in more than 50 countries. Survey respondents were asked where they worked via an open-ended question with an autofill option, and corporations were eligible for consideration if they employed more than 1,000 people and operated globally.
Get the full list: forbes.com/lists/worlds-best…
📸: Getty Images
Honoured to be working with you @MParakhin ! Very well deserved
Thank you @Time for the recognition time.com/collection/executiv… and congratulations to the full list of honorees!
Funny to see people say tool calling itself can be decomposed into classify + action. In 2010 that’s exactly how we built agents, and we dreamed of having one model combine NLU (intent and action classification and Entity extraction), dialogue management, and NLG into one. The world is a silly place.
jev is the best excuse you could possibly have to go re-read 12 factor agents. Tool calling itself can be decomposed into classify+action,
if you learn to design ai programs as pipelines that switch breathlessly between classification, structuring data, deterministic code, AND small agent-shaped append-chat loops, then jev is a WONDERFUL building block
hlyr.dev/12fa
I think I'm pretty good at detecting AI fluff, @appliedcompute is the real deal!
Working with @Drewch and the shopify team is awesome. Honestly never seen a group so customer obsessed and willing find the best path to delight them no matter what.
Plus an insane iteration speed by being so data + experiments driven!
I'm realizing from the flywheel thread from @tobi that many people don't realize that small models can actually outperform larger models on a specific task - @MParakhin has been saying this for quite some time. 🧵
Surely, there must be some downside to using a smaller model? Yes, there is, but the upsides outweigh the downside. You will lose generalizability, but you will vastly improve quality while reducing latency and cost.
Check out the talk at @icmlconf that @cmazzaanthony and I gave this summer. Stay tuned for @Shopify in December at @NeurIPSConf. icml.cc/virtual/2026/75732