About GenTube

GenTube is a consumer AI creation platform built on a simple belief: creation should be entertainment.

Last year, people created 70M+ images on GenTube. What matters more is what’s emerging now: a small but growing group opens the app with no prompt, no goal, and stays for hours. No nudges. No incentives. That behavior is the signal we’re building around.

We’re an early, opinionated team based in Toronto, backed by top consumer AI investors and operators who’ve built at global scale.

Our ambition is straightforward and hard: build the next great consumer AI creation company for a billion people.

The Role

We’re hiring a Product ML Engineer to build the intelligence layer of GenTube.

This is not a research-only role.

And not an infra-only role.

You’ll work at the intersection of models, systems, and product — shipping ML that real users feel every day. You’ll make explicit tradeoffs between speed, quality, cost, and delight — and measure them.

If you want ownership, rigor, and real-world scale, keep reading.

What You Will Do

Core ML Infrastructure

• Build inference pipelines serving millions of generations per week.
• Design real-time and streaming inference for diffusion models, LLMs, and multimodal systems.
• Optimize latency across serving, batching, caching, routing, and model selection.

Model Performance

• Adapt and productionize foundation models (SD, Flux, LLMs).
• Implement quantization, distillation, pruning, and compilation
• Experiment with LoRAs, ControlNets, adapters for style, control, and personalization.

Intelligence Layers

• Build ranking, recommendation, and personalization systems.
• Implement content understanding with embeddings, similarity search, clustering, classification.
• Build moderation and safety systems that scale without killing creativity.

Production Systems

• Scale GPU infrastructure from thousands to millions of daily generations.
• Profile bottlenecks and optimize utilization and cost.
• Run A/B tests on model variants; monitor quality, drift, and p99 latency.
• Own reliability, observability, and graceful degradation.

Relentless Experimentation

• Ship new model variants frequently.
• Test speed vs. quality tradeoffs using real user behavior.
• Close the loop: user behavior → signal → model improvement.

What We Are Looking For

ML Engineering Depth

• Proven experience shipping production ML systems at scale.
• Strong with generative models (diffusion, LLMs, multimodal).
• Fluency in PyTorch (JAX a plus).
• Experience with inference serving (Triton, Ray, TorchServe, custom stacks).

Systems & Infrastructure

• Deep knowledge of inference optimization: batching, quantization, compilation.
• GPU systems experience: CUDA awareness, memory management, multi-GPU serving.
• Strong backend skills: Python, async systems, APIs (FastAPI or similar).
• Comfortable owning systems in cloud environments (AWS/GCP/Azure).

Product-Focused ML

• You care about user experience, not just benchmarks.
• You design for sub-second latency, resilience, and delight.
• You can explain tradeoffs clearly and make the call.

Mindset

• Startup DNA: fast-moving, scrappy, high ownership.
• Performance-obsessed, quality-driven.
• Empathetic to creators — you understand intent, expectations, and failure modes.

Why Join

• Founders have scaled consumer products to 100M+ users and led a $150M+ AI exit.
• Backed by top consumer AI investors and operators.
• We’re building the kind of company Canada rarely builds — consumer-first, global, culturally relevant.
• Small team. High bar. No bureaucracy.
• A rag-tag group of pirates in the desert.

Location: Toronto (downtown). On-site.

Comp: Competitive salary + meaningful equity.

Benefits: Health, dental, vision, unlimited PTO, creative tools & education stipend.

Taste, curiosity, and ownership matter more than pedigree.

If you want to ship ML that millions of people feel, measure what works, and push the edge of consumer AI — we want to hear from you.

Apply by sending your application to [email protected]