← All roles

Backend Engineer

Channel3· Engineering
Apply on company site
New York, NY$120k - $200k + equityFull-timePosted 9mo ago

About the role

Channel3 is building a database of every product on the internet with a focus on product catalog and search infrastructure to enable agentic commerce. The Backend Engineer will work on understanding and indexing vast product data, building reliable and cost-efficient search, and reducing token usage across large-scale multimodal models.

Responsibilities

  • Write code to understand 1B products by stitching language, embeddings, image, and segmentation models to understand each product.
  • Deduplicate products across merchants using multimodal models.
  • Identify product variants, such as colors and sizes, and configurations for items (e.g., furniture).
  • Build world-class search enabling complex, fast queries with structured data and deterministic filters.
  • Engineer for reliability and cost: create evals to measure drift and guardrails to prevent regressions and hallucinations; reduce token usage and optimize models and data storage.

Qualifications

  • Any experience level accepted (new grads ok) - Visa US citizen/visa required.

Skills

Required:Backend engineeringMultimodal data processingSearch and indexing
Last updated Oct 21, 2026. Interested? Apply on the company site.
Channel3 logo

About Channel3

Database of every product on the internet

Channel3 is building a database of every product on the internet. Google and OpenAI maintain vast product graphs, but keep that valuable data to themselves. Channel3 gives every developer access, opening a world of possibility for new shopping experiences across any vertical. For instance, developers could build AI fashion stylists or give their agents shopping capabilities. We don’t just collect product data; we actually understand it. With the latest image classification and reasoning models, we can match products across merchants—even when listings and images differ—recognize variants, and surface the perfect matches for every search. As we grow, we’ll capture unique insights into what users want, click, and buy. As AI models get smarter and cheaper, our flywheel accelerates: our catalog will improve, and it will be even easier for developers and vibe-coders to build on our infrastructure. To further incentivize building on Channel3, we provide built-in monetization; developers earn commissions on sales they drive. Alex discovered this need while building an AI tutor that could recommend products and earn affiliate revenue. When no solution existed, he persuaded his smartest friend from Duke and former co-founder George to leave Palantir and build Channel3 together.

artificial-intelligencee-commerce