Product Engineer

About Etched

Etched is building AI chips that are hard-coded for individual model architectures. Our first product (Sohu) only supports transformers, but has an order of magnitude more throughput and lower latency than a B200. With Etched ASICs, you can build products that would be impossible with GPUs, like real-time video generation models and extremely deep & parallel chain-of-thought reasoning agents.

Job Summary

As Etched prepares to scale up customer deployments, we are seeking technical, product-focused members of the team who will bridge the gap between engineering and the go-to-market team. You will lead the design and implementation of our developer portal, which is the primary user-facing touchpoint for customers looking to deploy Sohu servers or developers looking to run their models on Sohu. This is a hands-on technical role that requires expertise in full-stack engineering, and you will collaborate cross-functionally with the software, go-to-market, and platform teams to architect the full customer interface to Sohu.

Key responsibilities

• Work with the Etched software team to define and document each layer of the Sohu software stack, from on-chip transaction execution to multi-rack workload sharding and management

• Implement live code execution within developer-facing materials

• Collaborate with designers to build performant and beautiful UI/UX for developers

• Connect interactive documentation with real production hardware to allow users to try out software workloads within sandboxed environments

• Curate and present demos of the emulation platform to customers; interface with customers to make and recommend use cases to engineering teams

• Work with the go-to-market team to uncover customer requirements and position Sohu capabilities to answer and inform customer questions

• Lead the design and implementation of the developer portal as the primary user-facing touchpoint for customers and developers

• Collaborate cross-functionally with software, go-to-market, and platform teams to architect the full customer interface to Sohu

You may be a good fit if you have

• Bachelor's degree or equivalent practical experience

• 5+ years of experience in software engineering, with a strong emphasis on full-stack development and systems design

• Demonstrated ability to build complex user-facing products with elegant UI/UX and a strong sense of developer ergonomics

• Proficiency with modern web frameworks (e.g., React, Next.js, or similar) and backend infrastructure (e.g., Python, Node.js)

• Familiarity with inference serving stacks (vLLM, SGLang), ML frameworks (e.g., PyTorch, TensorFlow), and AI hardware acceleration (e.g., CUDA, ROCm, other GPGPU paradigms)

• Technical depth: Deeply understands (or can quickly learn) how AI computing infrastructure works from an application layer, software stack, ML research, and data center perspective

• Proactive self-starter: Can work across teams to assemble materials, gather data, run meetings, and more with minimal assistance

• Opinionated: Spots problems, speaks up when disagreeing, takes ownership, and can handle making important decisions

Strong candidates may also have experience with

• Deploying AI applications and infrastructure in the cloud or with on-prem compute clusters

• Building developer platforms, portals, or technical documentation tools

• Working with APIs, containerization, CI/CD pipelines, and cloud infrastructure (e.g., Docker, Kubernetes, AWS/GCP)

• Infrastructure for distributed computing or datacenter-scale systems

• Supporting early-stage product launches, especially technical pre-sales

Benefits

• Full medical, dental, and vision packages, with generous premium coverage

• Housing subsidy of $2,000/month for those living within walking distance of the office

• Daily lunch and dinner in our office

• Relocation support for those moving to San Jose (Santana Row)

Compensation

• $175,000 - $225,000

How we’re different

Etched believes in the Bitter Lesson. We think most of the progress in the AI field has come from using more FLOPs to train and run models, and the best way to get more FLOPs is to build model-specific hardware. Larger and larger training runs encourage companies to consolidate around fewer model architectures, which creates a market for single-model ASICs.

We are a fully in-person team in San Jose (Santana Row), and greatly value engineering skills. We do not have boundaries between engineering and research, and we expect all of our technical staff to contribute to both as needed.