Voice Agent Engineer

Harrison Clarke; San Francisco (CA); On-site; USD 150,000 - 260,000; Full time

Job summary

Harrison Clarke, a well-funded early-stage AI startup in San Francisco, is seeking an experienced AI engineer to own the learning loop behind autonomous agents. You will help ship end-to-end AI products, from evaluation to deployment, directly impacting real-world workflows and human-AI collaboration.

Join founders to build production AI that learns from interactions, improves over time, and coordinates multi-step processes with speech and reasoning capabilities.

Qualifications

• Shipping LLM-based products into production.
• Building AI agents from the ground up.
• Experience with voice/chat interfaces.
• Experience with AI evaluations and benchmarking.
• Familiarity with RLHF or reward modelling.
• Ability to move quickly in high-ownership startup environments.

Responsibilities

• Own the complete learning loop behind autonomous agents.
• Develop evaluation frameworks and benchmarking infrastructure.
• Build reinforcement learning and feedback pipelines for production systems.
• Create continuous learning systems driven by real-world data.

Skills

Python

Reinforcement learning

RLHF

AI evaluation

Agentic AI

Startup ownership

Voice AI

Tools

LLM frameworks

Observability

Job description

We're partnered with a well-funded early-stage AI startup building a new generation of autonomous AI systems capable of handling complex, real-world workflows through voice, reasoning, and orchestration.

This isn't another AI wrapper.

The team is building production AI that communicates with humans, coordinates multi-step processes, makes decisions autonomously, and continuously improves from real-world interactions. Every customer conversation becomes training data for the next generation of the system.

As one of the earliest AI engineers, you'll own the complete learning loop behind these agents.

What you'll build

• Evaluation frameworks and benchmarking infrastructure
• Reinforcement learning and feedback pipelines
• Continuous learning systems driven by production data
• End-to-end AI products deployed to real users

You’ll likely have experience with

• Shipping LLM or agentic AI products into production
• Building AI agents from the ground up
• Voice AI or conversational AI
• AI evaluations, benchmarking, or observability
• RLHF, reward modelling, or feedback-driven learning systems
• Python and modern LLM frameworks
• Moving quickly in high-ownership startup environments
• Founding-level ownership from day one
• Work directly alongside experienced founders