Founding AI Engineer

• 🛗 Elevator pitch: JotPsych software listens to behavioral health providers while they talk to their patients (with consent, of course). A few seconds after the patient leaves the room, the provider already has a robust draft of the medical note waiting for them. Super customizable, easy to integrate with any current workflow. Medical note completely done and submitted for insurance reimbursement before the next patient enters the room. No more typing during the session, no more catching up on notes at home in the evening. Eye contact and peace of mind. 🧘

• 🚀 JotPsych is a post-revenue, seed-stage software startup backed by Base10. The product currently reduces administrative time for behavioral health clinicians by 90% and is expanding to encompass other clinical and administrative workflows. You can learn more about JotPsych by visiting the website, and/or by listening to the podcast below:

<https://open.spotify.com/episode/3N5UYm2tbBKdROlFcYcVlg?si=846cb7ebb28449e3>

• ⛓️ JotPsych was founded by Nate Peereboom and Jackson Bierfeldt. Nate’s mom was a nurse practitioner, so he grew up in the shadow of overwhelming medical documentation. Jackson is a language and audio nerd, and when GPT-3 hit the market, he knew something could be done here. The company is currently 5 FTE.

• 🚀 JotPsych has scribed for over 600,000+ patient encounters in its first 12 commercial months. Our revenue grows 10-30% every month, and we are rapidly approaching $1.5M ARR.

• Check outWho and how we hire

Role

We're looking for a Founding AI Engineer who can help us create and supervise agentic workflows that push the boundaries of what's possible in behavioral healthcare. We're focused on delivering real value to clinicians today through the automation of burdensome administrative work, and we’re building towards an entirely new category of software that improves behavioral health outcomes.

You'll be a critical part of our founding engineering team (member 3 or 4), working directly with our technical CTO. You'll be responsible for architecting, implementing, and orchestrating the AI systems that power our core product, focusing on building reliable, observable, and scalable solutions. Your mission will be to make AI feel natural, trustworthy, and useful to healthcare professionals.

That said, we’re an early-stage startup with a small team, so you’ll be working across the entire stack as well.

We believe in pragmatic AI development - we leverage the best available tools and APIs to solve real problems today. While we're doing innovative work, we're not a research lab (yet) - we're a product company focused on delivering immediate value to healthcare professionals.

Some early epics may include:

• Building robust evaluation frameworks to measure and improve our AI systems' performance

• Designing and implementing feedback loops that help our models learn and improve from real-world usage

• Optimizing our RAG pipelines for clinical workflows

• Designing AI Agent Workflows that automate high-complexity tasks

• Creating observable systems that help us understand and improve model behavior

• Implementing fine-tuning pipelines for specialized tasks

• Developing fallback systems and model redundancy for maximum reliability

• Building automated quality control systems for AI-generated medical documentation

Ideal Candidate

We're looking for someone who combines modern AI expertise with pragmatic engineering sense. You should have at least a few years of startup experience and have deployed a robust LLM-powered application to production. (Candidates just out of a master's program—we love you, but not for this role!)

You have hands-on experience bringing LLMs to production and understand the practical challenges of building reliable AI systems. You're excited about using AI to solve real problems for healthcare professionals, and you know that the hardest problems often aren't about the models themselves, but about making them work reliably and transparently in production.

• Strong software engineering foundation (5+ years), with recent focus on AI/ML systems

• Production experience with LLMs (OpenAI, Anthropic, etc.) and RAG architectures

• Experience building evaluation frameworks and feedback systems for AI

• Comfortable with rapid prototyping and iterative development

• Knowledge of LLM fine-tuning and prompt engineering best practices

• Experience with audio processing and ASR systems is a plus

• Strong focus on observability and system reliability

What stack experience do we hope you have? Here's what we currently work with:

• OpenAI, Anthropic, and other LLM APIs

• Assembly AI, Deepgram, and other ASR systems

• Vector databases and embedding models

• LangChain and similar LLM frameworks

• Python-based AI/ML tools and libraries

• AWS infrastructure for AI workloads

But let's be honest: if you're good, you can learn any of these in a short amount of time.

Timing and location

We want to hire as soon as possible. We aim to complete the hiring process in a few weeks.

🏡 🚞 This full-time role is officially remote, but with a strong preference for candidates based in the US and willing to travel somewhat frequently to our headquarters in Washington, DC. We have an office here that most of the team works out of.

Compensation

Compensation includes the following benefits:

• 🌎 Deeply meaningful work — you’re changing how behavioral health happens! Our customers literally hug us when they meet us at conferences and tell us how we’ve “changed their lives.” 🥲

• 💵 Cash (~$150-250k)

• 🧱 Material equity (~1.00%)

Challenge + Risks

The challenges we face are unique and exciting! We're not just implementing AI models - we're building systems that healthcare professionals trust with their most important work. Our users need to feel confident that our AI will help them create accurate, compliant medical documentation every single time. The biggest challenge isn't about using the latest models - it's about building reliable, observable systems that consistently deliver value. How do we ensure our models maintain medical accuracy? How do we build feedback loops that help us improve? How do we gracefully handle edge cases and failures? AI Agents are eating software. These are the problems you'll help us solve.

Application process

First, email us (engineers@smartscribe.health) to introduce yourself and explain why you want to work at JotPsych. Keep it short (200 words or less) but professional. Include your resume, LinkedIn, and portfolio / GitHub.

Ready for the M&M test? The subject of the email must include a pretzel emoji: 🥨. Because like a pretzel, AI systems have many delicious twists and turns

<aside> ⚠️ If you just “Easy apply” on LinkedIn or WellFound, we won’t look at your application.

</aside>