T

Town.com

AI personal assistant automating email and scheduling

AI Engineer - Evals & Agent Quality

Natural Language Processing & LLMsConfirmed live in the last 24 hours
Full-timeSan Francisco, CA, USA$250k - $300k /yr

About the Job

Requirements

  • Have built or owned large language model evaluation systems, or offline/online quality measurement at scale.
  • Think rigorously about measurement; an applied machine learning or machine learning engineering background is one possible source, but exact pedigree is not required.
  • Know evaluation tooling and frameworks hands-on, including off-the-shelf tools, and have opinions on which to use.
  • Be comfortable reasoning about model routing and tradeoffs between models.
  • Be a senior or staff engineer comfortable working in a greenfield environment.

Responsibilities

  • Build a generalized evaluation system that measures assistant quality across every surface it touches, including multi-step agent trajectories.
  • Establish golden datasets and an ongoing labeling loop to keep them current and continuously validate improvements and avoid regressions.
  • Build model routing and online evaluation tooling to help the team learn and route to the best models.
  • Make every prompt and system change measurable so the team can move quickly without breaking what works.
  • Partner with engineers across the product to instrument quality and close the loop from signal to fix.
  • Ship fixes, not just dashboards and metrics.

About the Company

Town.com

AI personal assistant automating email and scheduling

Town builds an AI work assistant called Townies that learns how an individual works and handles email, scheduling, meeting prep, and documents across connected tools like Gmail, Google Calendar, Slack, WhatsApp, Telegram, and Notion. Each user gets a personal assistant that builds a knowledge base of people, projects, and voice, drafts replies, triages inboxes, prepares meeting briefings, and runs recurring routines. It includes permission controls about what it can send or change and by default does not use data to train models. It serves individuals and teams through subscription plans from free to Team tiers.

Company size
51-200
Company stage
Series B
Total funding
$181M
Headquarters
San Francisco, CA
Founded
2024

Recently Posted Jobs

Explore more opportunities at Town.com

See more jobs