Founding Engineer - On-site New York, NY - $150K - $250K

What one of clients is looking for
We need a builder with strong full-stack or backend engineering skills and deep hands-on experience with applied AI. That means real agent harnesses where you understand tool calls, memory, system prompts and retrieval, not just thin wrappers around an OpenAI SDK. You should be comfortable owning whatever lands on your plate in a small, fast, in-person team in NYC. You should also have a track record of launching your own projects that produced real results, like users acquired or revenue generated. Bonus points if you have strong consumer product sense, founding engineer experience at an applied AI startup, open-source work, or a long GitHub commit history from before AI coding tools.
What you'll do:
Build and improve our client's core agent harness: tool use, memory, system prompts, retrieval and evaluations. This is the top priority from day one.
Own features across the consumer chat app and backend from architecture through launch and iteration, for 500K–700K daily active users and 5M+ total users.
Work directly with the founders to merge chat, agents, tools and the growing app library into one simple, fun consumer AI experience.
Expand agent capabilities by giving agents VMs and more advanced tooling, plus backend and database infrastructure that can scale to 100M+ users.
Help deploy and optimize roughly $5M of incoming B300 GPUs, including fine-tuning and serving efficient open-source models on sustainable data center infrastructure.
Balance latency, cost, reliability, privacy and user experience, and contribute to future projects like the EcoGPT phone and an AI-native Android OS.
Requirements
Dealbreaker
Work experience
Has built and shipped their own product (startup,​ side project,​ or open source) with measurable results:​ users,​ revenue,​ or adoption
Hard skills
Hands-​on with open-​source LLMs:​ serving,​ fine-​tuning,​ or evals
Soft skills
Strong communication
Miscellaneous
Public GitHub with personal projects and consistent,​ recent commit history
Baseline
Seniority
1 -​ 5 years of experience 1+ years of hands-​on engineering,​ either as a SWE at a startup,​ or through sustained independent/open-​source projects (coursework doesn’t count)
Work experience
Has built an agent harness or agentic system hands-​on (tool calling,​ memory,​ system prompts),​ beyond calling an LLM SDK
Miscellaneous
Able to work up to 7 days a week (5 in office)
Nice-to-have
Work experience
Founding or early engineer at an applied-​AI or consumer startup
Tech stack
AI Agents / Agent Harnesses, LLM Tool Calling, Agent Memory Systems, Prompt Engineering / System Prompts, Retrieval (RAG), LLM Evaluations, Open-Source LLMs, Model Fine-Tuning, GPU Infrastructure (NVIDIA B300), Agent VMs / Sandboxed Tooling, Full-Stack Development, Backend Engineering, Scalable Databases, Android OS (AI-native)
Salary: $150K - $250K Equity: 1%-3%
Visa sponsorship details
Open to visa transfers (e.g. OPT, H1B transfers)
Additional visa sponsorship details
Canadians on TN visas fine; Other visa transfers possible for exceptional candidates.
On-site work policy
Fully in person at the New York City office. The team is in the office most days, often 7 days a week. Hours aren't fixed and the culture is focused on output, but candidates should expect an intense, all-in early-stage pace. There is a strong preference for people who can work in person, with only slight flexibility for exceptional engineers. No relocation support. Sourcing priority: New York, San Francisco and Canada (especially Waterloo). The interview process includes a paid remote trial of about 1–2 weeks (flexible), followed by a few days working on-site in NYC.
Full-time position
Location
New York, NY

Work arrangement
Yes