Member of Technical Staff
The Token Company
Member of Technical Staff
The Token Company
AI middleware for semantic prompt compression
The Token Company (YC W26, HF0 S26) is a seed stage startup in San Francisco, CA, and has raised ~$12M from First Round Capital, NEA, YC and SV Angel along with investors such as founders of Dropbox, Slack, Supercell and Huggingface and key people from OpenAI, xAI and DoorDash.
The Token Company trains machine learning models to compress raw LLM inputs before they are passed to the expensive LLMs. Our models are able to recognize parts of inputs that are unimportant, reducing inference costs for scale-ups and enterprises that integrate LLMs into their products. The team is 5 people with a research and product focus.
As a Member of Technical Staff on our infrastructure team, you'll own the cloud systems that serve our compression API end-to-end. You'd get to build global low-latency, high-throughput GPU ML inference infra that sits in the critical path of customer traffic, from deployment and scaling to reliability and cost-efficiency. This is a very high ownership role where you'd need to learn a lot of new solutions and technologies, constantly improving and researching what's best. The ideal candidate fits our culture and is:
A quick learner: grasping the product and the systems around it and moving fast is important
Strong in cloud infrastructure: comfortable with AWS, Terraform, Docker and CI/CD, and building for performance and reliability at scale
Have owned an infra stack end-to-end: have built and operated production infrastructure yourself before, at a startup or a larger company
In addition to base salary, this role is eligible for significant equity. We also provide housing, food, laundry and cleaning in SF, visa sponsorship, and a position to build the team around you with our resources.
Our website: thetokencompany.com
AI middleware for semantic prompt compression
Token Company (Otsofy) offers an AI middleware API that semantically compresses LLM prompts to reduce token usage, latency, and costs. It uses a fast model called bear-1 to preprocess inputs before sending them to LLMs like GPT or Claude, preserving meaning while removing nonessential tokens. The API is a drop-in integration for production use and aims to be a neutral efficiency layer across models, helping teams run larger prompts within fixed compute budgets. Its goal is to become a core part of the AI stack by improving efficiency for every LLM request.
Explore more opportunities at The Token Company