Production AI Engineer: Ship LLM Features & Agents

Digital Waffle; San Francisco (CA); On-site; USD 170,000 - 250,000; Full time

Job summary

Digital Waffle in San Francisco is hiring an AI Engineer to design, build, and ship LLM-powered features and agentic workflows in production. You’ll work close to product and users with an applied AI stack—prompts, retrieval, tools, agents, evals, and the glue to our product.

You’ll design production LLM features using Python/TypeScript, build end-to-end tools like function calling and RAG, and integrate with model APIs. 5 days in office, strong backend/API foundation desired.

Qualifications

• Production experience shipping LLM-powered features.
• Strong Python and TypeScript skills with production readiness.
• Experience with function calling, RAG, vector databases, structured output, prompt engineering.
• Familiarity with an agent framework/SDK and its abstractions.
• Experience building evals and guardrails for LLM outputs.
• Solid API, integration and backend engineering fundamentals.

Responsibilities

• Design and ship production LLM features and agents using Python/TypeScript.
• Build the applied AI toolkit end-to-end: tools, function calling, RAG, vector search, structured output, prompts.
• Work with foundation model APIs (Anthropic, OpenAI, Google) and open-source models.
• Create evaluation pipelines to measure quality.
• Integrate agents into product surfaces built with React/Next.js.
• Own cost, latency and reliability of the AI layer.

Skills

Python

TypeScript

LLM features

Agent framework

Evaluations/guardrails

API/backend

Vector databases

Tools

OpenAI API

LangSmith

Claude/Cursor

Job description

Digital Waffle in San Francisco is hiring an AI Engineer to design, build, and ship LLM-powered features and agentic workflows in production. You’ll work close to product and users with an applied AI stack—prompts, retrieval, tools, agents, evals, and the glue to our product.

You’ll design production LLM features using Python/TypeScript, build end-to-end tools like function calling and RAG, and integrate with model APIs. 5 days in office, strong backend/API foundation desired.