Logo
Qumis

Founding Staff AI Engineer

Qumis, New York, New York, us, 10261

Save Job

Founding Staff AI Engineer Qumis

is a venture‑backed AI startup building an advanced platform for reading and interpreting insurance policies.

We help P&C insurance professionals—claims advocates, brokers, underwriters, and legal teams—move faster, make better decisions, and reduce reliance on siloed expertise.

Our vision is a world where anyone can read any insurance policy like a lawyer.

Base pay range $150,000 - $250,000 per year

About the Role You’ll partner closely with the CTO and co‑founders to design, implement, and own every part of our AI pipeline—from unstructured document ingestion to interactive LLM outputs that shape legal‑grade decisions.

It’s a product‑critical role for someone who thrives in ambiguity, ships fast, and is excited to be the technical owner of a new AI stack.

What You’ll Do

Build and ship production AI features end‑to‑end—from R&D to SaaS integration, observability, and iteration

Design and deploy advanced document understanding systems, including RAG pipelines, clause segmentation, summarization, risk detection, and more

Create scalable APIs that deliver real‑time coverage analysis, comparisons, and recommendations inside a modern SaaS UI

Engineer robust data workflows across OCR, PDF ingestion, parsing, normalization, and embedding generation

Own evaluation and observability, from automated accuracy pipelines to human‑in‑the‑loop systems that maintain trust and reliability

Design for explainability—develop transparent workflows that help legal, claims, and compliance professionals understand what the AI sees and says

Collaborate across disciplines with designers, PMs, and domain experts to translate insurance workflows into AI‑powered systems

Set technical direction—you’ll shape our infra, tooling, and AI product roadmap from the ground up

What We’re Looking For

6–10+ years of experience building and deploying production machine learning systems, with a focus on NLP, document intelligence, or unstructured data

Strong track record shipping ML models into production—API design, observability, iteration, and scaling included

Deep experience with RAG pipelines, multi‑step LLM workflows, and vector databases (e.g., Pinecone, Weaviate, Qdrant)

Comfortable working with complex, messy documents (e.g., insurance policies, contracts), including parsing, OCR, clause‑level segmentation, and normalization

Strong software engineering background, with experience in full‑stack dev, containerization (Docker), and DevOps/MLOps best practices

Ability to design secure, compliant systems in regulated environments (insurance, legal, healthcare, or finance)

Experience architecting agentic workflows and AI‑native products that go beyond simple chatbot interfaces

Product‑first mindset—you build models to solve real user pain, and love working closely with cross‑functional teams

Operates like a founder—driven, hands‑on, scrappy, and thrives in fast‑paced, zero‑to‑one environments with extreme ownership

Why Join Qumis

Backed by top‑tier venture capital funds and strategic angels in AI and insurance

Used and loved by leading brokerages and claims teams

Shape the future of generative AI in one of the world’s most complex industries

High‑ownership role on a high‑standards founding team with grit, urgency, and big ambition

Build a generational company—an opportunity that only comes once in a career

Benefits

Medical, dental, and vision insurance (Aetna + Guardian) with 100 % premium coverage for employees and up to 90 % for dependents

Life insurance (company‑paid), supplemental options, short‑ and long‑term disability insurance, pre‑tax FSA and HSA options, pre‑tax commuter benefits

Unlimited PTO & $1,000 home office setup stipend

Hybrid work environment: 1–2 days/week in our Chicago office

Professional development support and growth opportunities

Seniority level Mid‑Senior level

Employment type Full‑time

Job function Engineering and Information Technology

Industries Software Development

#J-18808-Ljbffr