Founding AI Full Stack Software Engineer (NYC Onsite) 200K+

Averity; New York (NY); On-site; USD 140,000 - 210,000; Full time

Job summary

Averity is expanding its engineering team and seeking an experienced, generalist engineer who can own deliverables end-to-end across the stack. The role emphasizes building AI-native features and agentic workflows, with autonomy to discuss problems, propose features, and ship production-ready software.

You will work on data extraction from EHR platforms, analytics, and scalable infrastructure while collaborating with therapists and practice owners as needed.

Qualifications

• Extensive professional experience with Claude Code, Codex, or both.
• Ability to own roadmap and execute end-to-end across the stack with limited guidance.
• Experience building AI-native features and agentic workflows with LLMs.
• Strong TypeScript and React experience.
• Experience operating production systems on AWS.
• Strong PostgreSQL and data-system fundamentals.

Responsibilities

• Pull data from EHR platforms and transform it reliably at scale.
• Build and improve analytics, metrics, reports, and operational insights.
• Create features such as prioritized tasks and recommended actions.
• Build "talk to your data" and other AI-native experiences.
• Own production features that rely on LLMs and agentic workflows.
• Deepen and expand our EHR integrations.

Skills

TypeScript

React

AWS

PostgreSQL

LLM integration

Strong communication

Tools

FastAPI

ECS

Playwright

GitHub

RDS

ElastiCache

EC2

Job description

Our company builds software and operational infrastructure for psychotherapy practices. We help therapists and practice owners understand how their businesses are performing, identify what needs attention, and reduce the administrative work required to run a practice. We're expanding our engineering team as the company grows.

The Role

We're looking for an experienced, generalist engineer — someone who can own a deliverable

end-to-end, across the stack, without needing well-defined requirements or close management.

We're also building toward a product where AI sits at the core of how the system works, not

bolted on after the fact — so experience building agentic or LLM-powered features is a strong

You’ll own both the roadmap and the execution for whatever you take on — discussing

problems, identifying opportunities, forming opinions, proposing features, making tradeoffs,

building the product, deploying it, and operating it in production.

What You’ll Work On

Depending on where you land, your work may include:

• Pulling data from EHR platforms and transforming it reliably at scale
• Building and improving analytics, metrics, reports, and operational insights
• Creating features such as prioritized tasks and recommended actions
• Building "talk to your data" and other conversational, AI-native experiences
• Building production features that rely on LLMs and agentic workflows
• Deepening and expanding our EHR integrations
• Improving reliability, scalability, and data quality across the platform
• Owning the AWS infrastructure behind whatever you build
• You’ll occasionally speak directly with therapists and practice owners, but this is primarily a
• product-building role rather than a customer-success role.

How We Use LLMs

We use Claude Code and Codex as core engineering tools, and we're looking for someone with

a sophisticated, high-throughput agentic workflow, not occasional AI-assistant use. That might

mean running many agent sessions in parallel, splitting large initiatives into independent

workstreams, or building tooling around agent orchestration, review, and testing.

At the same time, we don't want someone who relies on LLMs for basic reasoning. You should

have developed strong engineering judgment before coding agents were widely available,

understanding how systems work, catching bad assumptions, and navigating hard technical

decisions independently. We're looking for someone who was already an excellent engineer and

has used LLMs to become dramatically more effective.

What We’re Looking For

• Comfort working from loosely defined requirements without close management — this is
• not a role for someone who needs to be managed, or a junior engineer
• Extensive professional experience with Claude Code, Codex, or both
• Experience coordinating complex work across multiple coding-agent sessions
• A track record of owning and shipping substantial production products
• Strong TypeScript and React experience
• Experience operating production systems on AWS
• Strong PostgreSQL and data-system fundamentals
• Experience building production features that use LLMs, ideally agentic ones
• The ability to work across product, backend, frontend, data, infrastructure, and
• operations
• Strong written and verbal communication
• The ability to disagree directly and constructively
• Empathy for users and teammates
• Comfort operating with limited guidance

Our stack: FastAPI, React, PostgreSQL, ECS, RDS, ElastiCache, EC2, Playwright, and GitHub

Actions.

Pluses Include:

agent tooling is useful but not required.

This is an individual-contributor role with real scope and influence over the product. It may grow

into technical leadership over time, but that's not a requirement to take it on.

How We Work

• We're an early-stage company with aggressive goals, and the pace reflects that — this is a
• high-intensity environment, not a standard 9-to-5. At the same time, you'll have substantial control over how you structure your day. We care about outcomes, communication, and ownership rather than fixed hours or performative availability.
• This is a high-autonomy position for someone who wants significant responsibility and influence.
• You'll work directly with our founder and foundational engineer, alongside an engineer focused on data and integrations