Founding Engineer

Wintermeyer Ventures; San Francisco (CA); On-site; USD 180,000 - 280,000; Full time

Job summary

Wintermeyer Ventures in San Francisco, California, USA is seeking a founding engineer to own large parts of our product end-to-end, working at the intersection of LLMs, unstructured data, and complex enterprise workflows where getting the answer right matters.

You will build AI agents and backend features from scratch, ship production capabilities from discovery through deployment, and work on-site with enterprise customers to deeply understand their workflows, data, and pain points.

Qualifications

• 2+ years of experience building and shipping production LLM or AI agent systems.
• Strong proficiency in TypeScript and backend development.
• Hands-on prompt engineering, eval design, and agentic workflow design.
• Background at an early-stage or high-growth startup, with shipping real work.
• Active use of agentic engineering tools as part of workflow.
• Comfortable working directly with enterprise customers and occasional on-site travel.

Responsibilities

• Build new AI agents and product features from scratch, owning entire modules from customer discovery through production deployment.
• Work on-site with enterprise customers to deeply understand their workflows, unstructured data, and pain points.
• Design and iterate on eval frameworks to ensure AI agents produce accurate, reliable outputs in specialized regulated domains.
• Build backends that ingest, structure, and reason over large volumes of unstructured data using LLMs.
• Act as a full-stack problem solver: writing code, conducting user research, self-educating on domain topics, and doing whatever it takes to make customers successful.

Skills

TypeScript

Backend development

Prompt engineering

Eval design

Agentic engineering

Enterprise software

Education

CS degree or STEM equivalent

Job description

About the Role

This is a founding engineer role at an early-stage AI startup building intelligent agent platforms for top-tier insurance carriers. You will own large swaths of the product end-to-end, working at the intersection of LLMs, unstructured data, and complex enterprise workflows where getting the answer right really matters.

What You'll Do

• Build new AI agents and product features from scratch, owning entire modules from customer discovery through production deployment.
• Work on-site with enterprise customers to deeply understand their workflows, unstructured data, and pain points.
• Design and iterate on eval frameworks to ensure AI agents produce accurate, reliable outputs in specialized regulated domains.
• Build backends that ingest, structure, and reason over large volumes of unstructured data using LLMs.
• Act as a full-stack problem solver: writing code, conducting user research, self-educating on domain topics, and doing whatever it takes to make customers successful.

What We're Looking For

• 2+ years of experience building and shipping production LLM or AI agent systems.
• Strong proficiency in TypeScript and modern backend development.
• Hands-on experience with prompt engineering, eval design, context construction, and agentic workflow design.
• Background at an early-stage or high-growth startup, with clear evidence of shipping something real and genuinely hard.
• Demonstrated standout achievement: competitive athletics, academic competitions, founding or joining a startup early, open source with real users, or equivalent.
• Actively uses agentic engineering tools (such as AI-assisted coding environments) as a core part of your workflow.
• Comfortable working directly with enterprise customers and open to occasional on-site travel.
• Experience in regulated or data-intensive domains such as insurance, fintech, healthtech, or legaltech is a strong plus.
• CS degree or STEM equivalent preferred.

Compensation & Benefits

Base salary range: $180,000 to $280,000 USD annually, plus equity. Visa sponsorship is not available.

Location

On-site in San Francisco, California, USA. Occasional travel to customer sites is expected.