TRECQ

AI Trainer (Part-time)

OtherSeniorremoteRemote
FullTimePosted Mar 10, 2026

About this role

About Kastle

Kastle is building the AI operating system for consumer lending. We partner with some of America’s largest mortgage lenders to transform their contact centers and compliance operations with AI voice agents.

Backed by Y Combinator, Commerce Ventures, and executives from Snapdocs, Google, and WePay, we’re redefining how millions of borrowers interact with financial institutions.

This is a rare chance to join at the ground floor of a company already plugged into large-scale customers and real production use cases.

Role Overview

We are looking for a part-time AI Trainer with experience operating phones in the mortgage industry. You will help improve our AI system by providing feedback on conversations, annotating data, and testing for quality. If you have worked as a customer service or collections agent for a mortgage servicer, this role is a great fit to use your knowledge to train AI in real-world scenarios.

Responsibilities

  • Annotate and label data from mortgage-related conversations.

  • Give feedback on AI responses to make them more accurate and helpful.

  • Train the AI system to handle mortgage servicing phone calls better.

  • Test the AI for issues, including response quality, customer experience, explanation quality, and realism.

  • Work ~10 hours per week on flexible schedules.

Requirements

  • 5+ years a customer service agent for a mortgage servicing company.

  • Knowledge of the mortgage industry, including escrow, payment questions, loss mitigation calls, QRPC requirements, servicing transfer questions, and customer conversations.

  • Strong attention to detail and communication skills.

  • Reliable internet and computer for remote work.

Preferred Qualifications

  • 3+ years as a Mortgage Customer Service Supervisor.

  • Familiarity with giving feedback to customer service agents.

Benefits

  • Flexible part-time hours.

  • Opportunity to shape cutting-edge AI technology.

  • Work from home.

About Kastle

AI employee for consumer lending

15 employeesYC Summer 2024
B2B

Interested?

Find the right person, then reach out. Trying more than one channel only helps your odds.

1
Find who to reach
Pick the person worth a direct note. Reveal a real email (one credit) or find a likely one.
New accounts start with 3 free reveals.
2
Reach out
Pick someone above, then draft an email or a LinkedIn note to them.
3
Or apply on their site
The front-door route, plus TRECQ drafts your application answers to paste.
Apply on Kastle ↗
TRECQ writes your "why this company / why you" to paste into the form.