Founding AI Engineer

True Talent Solutions Partners•New York, NY
•Onsite

About The Position

Join an early-stage SaaS team as its fifth engineer and lead the technical direction of a real-time AI voice agent for sales. The agent participates in live customer calls, demonstrates products through browser control, and supports follow-up through the sales process. Build the agent's reasoning, tool use, memory, and ability to recover when live conversations go off script while meeting strict latency requirements. Develop evaluation pipelines and data workflows that turn real sales calls into measurable product improvements, using fine-tuning where appropriate. Improve the reliability of browser-based product demonstrations conducted live while the agent speaks with buyers. Ship end-to-end product features using TypeScript, Node.js, and Next.js, working directly with customers and design partners. Set technical direction for model and architecture choices, build-versus-buy decisions, and future agent capabilities. The position is on-site in New York, United States. Base salary is $216,000 to $250,000 annually. Visa sponsorship is available.

Requirements

  • Typically 4 to 7 years of experience as an AI engineer in voice AI, AI sales, browser agents, or fast-growing AI companies.
  • At least one year owning a user-facing AI agent in production, with real users or meaningful usage volume.
  • Experience building and operating AI agents end to end, including evaluation, observability, and cost optimization.
  • At least one year working with real-time voice technologies such as WebRTC, LiveKit, or streaming systems, or browser agents using tools such as Playwright.
  • Proficiency in TypeScript, Node.js, and Next.js, along with experience with LLM APIs, browser automation, and agent evaluations.
  • Customer-facing engineering experience, including working directly with design partners or customers.
  • Ability to work on-site in New York from the start.

Nice To Haves

  • Sales technology experience is useful.

Responsibilities

  • Build the agent's reasoning, tool use, memory, and ability to recover when live conversations go off script while meeting strict latency requirements.
  • Develop evaluation pipelines and data workflows that turn real sales calls into measurable product improvements, using fine-tuning where appropriate.
  • Improve the reliability of browser-based product demonstrations conducted live while the agent speaks with buyers.
  • Ship end-to-end product features using TypeScript, Node.js, and Next.js, working directly with customers and design partners.
  • Set technical direction for model and architecture choices, build-versus-buy decisions, and future agent capabilities.

Benefits

  • Visa sponsorship is available.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service