Founding AI Engineer at Pocket
**About the Company**
At Pocket, we’re on a mission to transform the rehab therapy space with end-
to-end, vertically integrated AI solutions. We’re backed by 2048 Ventures and
our first product, insurance verification, is live with rehab therapy clinics
processing thousands of verifications per month. This year, we’re rolling out
a suite of new capabilities to streamline how rehab therapy clinics operate
using LLM-based text and voice systems.
**About the Role**
You will report to and work closely with the CTO, who has 8 years of
experience developing text and voice AI products for both enterprise and
early-stage startups.
We are looking for senior candidates who have operated in fast-paced, early-
stage startup environments and exhibit strong engineering judgment around
speed, reliability, and complexity trade-offs.
In this role, you’ll own the end-to-end development lifecycle, including data,
evaluation, deployment, and reliability. We’re looking for candidates who
thrive in high-trust, high-ownership environments.
**Key Responsibilities**
Build speech training & evaluation frameworks using prompting, fine-tuning,
and retrieval-based techniques
Implement document generation training & evaluation frameworks for clinical
workflows
Develop test harnesses and monitoring systems for critical AI workflows
Continuously refine datasets and evaluation methodologies to improve model
performance and real-world alignment
Build automated pipelines for training, deployment, and monitoring of AI
systems
**What We’re Looking For**
5+ years of professional experience in applied ML, data science, or software
engineering with at least 2 years in a startup environment
Experience developing training & evaluation frameworks for LLM-based speech
processing, including ASR, diarization, TTS, and speech-to-speech systems
Experience deploying AI speech solutions in production settings using REST
API’s, containerization and CI/CD pipelines
Strong Python expertise
**Compensation & Benefits**
$150K base salary
0.75-1.0% equity (stock options) fully diluted at time of grant, vesting over
4 years with a 1 year cliff
Flexible PTO
Health insurance
**Location**
This is a remote position and the company is headquartered in New York City.
**Our Hiring Process**
Initial Screen (30 minutes)
Technical Interview (1 hour)
Systems Design & Behavioral Interview (1 hour)
Meet with CEO (30 minutes)
Meet with both co-founders (30 minutes)
**First 60 Days**
Deliver monitoring systems and evaluation harnesses for insurance voice and
clinical documentation generation workflows
Drive a measurable improvement in the AI voice systems through targeted error
analysis, data iteration, and model optimization
Ship 1-2 production voice agents for real-time insurance payer interaction
Improve the documentation generation system through prompt and fine-tuning
based techniques
Collaborate with the CTO to deliver additional integrations and enhancements
to the platform