Help Define How Mental Health Is Assessed and Measured

×

Engineering Team Lead

Engineering, Boston area preferred

The role

Deliberate AI is hiring an Engineering Team Lead: a senior, hands-on engineer who sets the standard for how a small team builds regulated, AI-enabled software and the AI agents inside it, with coding agents in the loop.

You will be the technical lead for the platform that delivers Deliberate’s AI to clinicians and study participants: the study administration API and web console, native mobile apps, and a real-time voice conversation agent. You will own delivery and technical quality, and you will be the engineer the rest of the team learns from.

This is a hands-on technical leadership role with no people-management duties. You write and review code, make the architecture calls, and raise the engineers around you.

What you will do

• Provide technical leadership. Own the technical direction and the quality bar, and drive technical delivery in partnership with product and data science.
• Build production AI agents. Design and engineer our conversational agents: tool and MCP integration, prompt and context management, evaluation, and safe failure handling, working with data science on behavior and with the platform on reliability.
• Set the standard for agentic coding. Define how engineers and coding agents work together: repo-level agent instructions (AGENTS.md), version-pinned agent skills, automated agent review on every pull request, and clear limits on what an agent may change. Agent output is treated as untrusted until tests and a human reviewer confirm it; every agent-authored change carries provenance; a human reviews and merges.
• Mentor junior engineers. Use code review, pairing and design review to teach fundamentals: testing strategy, API and schema design, safe database migrations, observability, security. Teach them to specify, supervise and verify agent work, not only to prompt it.
• Define clean, effective architecture. Keep service boundaries and contracts explicit across the API, MCP servers and voice agent (contract tests, versioned schemas). Record significant decisions as ADRs.
• Run delivery under change control. Our software is regulated (Class B). Work is traceable in Jira, reviewed under documented SOPs, and infrastructure changes that touch security controls carry a risk assessment and verification evidence.
• Evolve the platform. Lead the move to isolated, private-by-default dev, staging and production environments on GCP, managed in Terraform.
• Partner with data science. Turn models, prompts and agent behavior into reliable production services with clear interfaces and ownership.
• Stay hands-on. Ship code and review code every week.

What you bring

Required

• 10+ years building and operating production software, including mentoring other engineers.
• Strong software engineering fundamentals that you can teach: automated testing, API and data-model design, migrations, CI/CD, observability, secure coding.
• Production experience in TypeScript on Node and Python, and with PostgreSQL.
• Experience building LLM-based agents for production use: tool calling, MCP integration, context management, and handling model failure safely.
• Daily, hands-on use of coding agents (Claude Code or equivalent) on real systems, and a clear view of their failure modes and the guardrails that contain them.
• Cloud and infrastructure-as-code experience; GCP and Terraform preferred.
• A record of raising the bar on a team: review culture, written design decisions, junior engineers who grew under you.

Preferred

• Evaluation of LLM and agent behavior: test harnesses, regression suites, and quality gates for prompts and tools.
• Real-time voice or WebRTC systems (we use LiveKit).
• Model Context Protocol (MCP) servers and consumer-driven contract testing (we use Pact).
• Working knowledge of native iOS or Android delivery.
• Experience in medical-device or other regulated software (for example IEC 62304, HIPAA) and change-controlled release processes.

Apply for this role