Product Engineer (Mid Level to Senior)

Teal HQ•Miami, FL

7d•Remote

About The Position

Teal is building the future of career infrastructure with AI-native agents that scout opportunities, prep for interviews, negotiate, and manage career data. This role involves joining a small, senior engineering team to rebuild the core experience around these agents, shipping features directly to millions of users. The focus is on developing AI-native solutions for complex, unsolved problems in career management, such as contextual evaluation, reasoning over messy data, efficient multi-step reasoning, and building a compounding career memory. The engineer will own AI features end-to-end, from API and model orchestration to UI and evals, working closely with the CEO and users.

Requirements

Shipped real LLM features in production, including tool use, RAG, and agents that don't silently fail.
Experience with multi-step agents and opinions on orchestration patterns.
Full-stack capability with strength in React/TypeScript or Rails, and comfort in both.
Willingness to work across the full stack without hand-offs.
Taste in evals, understanding the pain of not having them and having a point of view on good ones.
Experience using AI tools like Claude Code, Cursor, Codex, OpenCode, and understanding their velocity benefits.
Strong sense of craft, caring about UX, speed, and polish.
Bias for closure; ability to take a fuzzy problem and deliver a working solution.
Desire for customer proximity: willingness to talk to users, read tickets, and sit on sales calls.
US-based with no sponsorship available.

Nice To Haves

Vector databases in production
MCP servers
Experience with eval frameworks (Braintrust, Langfuse, homegrown)
Experience in careers / HR / recruiting products

Responsibilities

Ship AI features end-to-end, owning the API, model orchestration, streaming UI, evals, and rollout.
Develop agentic workflows, including multi-step agents for role research, career analysis, bullet writing, interview prep, and offer negotiation.
Build streaming interfaces that are instant, handle retries gracefully, and degrade well.
Create and maintain the eval system, including harnesses, regression suites, and latency/cost budgets.
Develop prompt and feature iteration infrastructure for A/B testing prompts without shipping releases.
Engage in direct customer interaction through calls, support tickets, and user interviews.
Contribute to architectural decisions regarding agent state, retrieval, model orchestration, and evals.
Work with edge-first AI infrastructure on Cloudflare (Workers, Durable Objects, AI Gateway, Vectorize, Queues, R2).
Implement multi-model agent orchestration (Anthropic, OpenAI, etc.) with tool calling, structured outputs, and streaming responses.
Develop the interface layer using React, TypeScript, shadcn/ui, and Tailwind, focusing on streaming and agent responsiveness.
Build and enhance the career memory (vault) using Postgres and pgvector for semantic recall across years of work.
Maintain and improve existing backend systems (Ruby on Rails, Python/FastAPI).