How AI Supports Doximity Teams and the Physicians Who Use Its Platform
AI is transforming Doximity — inside and out.
According to Senior Software Engineer I Ronnie Rocha, AI shows up in the solutions Doximity engineers ship and in nearly every workflow related to product development.
But AI isn’t limited to the engineering team. It also enhances workflows across the organization, as QA employees, data team members and designers share insights and lean into various AI tools to drive efficiency in their day-to-day work.
“Nobody mandated it,” Rocha said. “People just saw their coworkers getting real results and wanted in.”
What Does Doximity Do?
Doximity’s digital platform helps physicians collaborate with colleagues, stay up to date with the latest medical news and research, manage their careers and on-call schedules, streamline documentation and administrative paperwork, and conduct virtual patient visits.
Doximity leaders see AI as an empowering tool that frees up time for people to focus on high-level strategy rather than tedious, repetitive work. And perhaps what’s most exciting for engineers like Rocha is the ownership he and his teammates have in deciding what AI tools get adopted.
“I’m given the freedom to experiment and refine my own workflows,” Rocha said.
Below, Rocha shares more about Doximity’s approach to leveraging AI — and the culture of autonomy and experimentation that makes the healthtech company an exciting place to work for tech talent.
How Doximity Uses AI in Physician Products and Team Workflows
Explain how AI is used at Doximity. How is it connected to your product or product development?
AI shows up in two places at Doximity: in what we ship to doctors and in how we build it. On the product side, we have Doximity Ask, which helps physicians handle things like drafting letters and looking up clinical evidence, and Scribe, which listens during patient visits and turns the conversation into structured notes. I was on the iOS team that built both of these into our mobile app.
On the development side, AI has worked its way into most of our workflows. One of our product managers started a Slack channel called “Get Stuff Done with AI,” where people across the company swap tips on what’s actually useful. PMs run Claude Code and Cursor. On our iOS team, we built a custom skill that turns a plain English description into a formatted Jira ticket, pulling in Figma design specifications and GraphQL schemas on its own. Engineers, QA employees, data team members and designers have all started picking these tools up. Nobody mandated it. People just saw their coworkers getting real results and wanted in.
What Are Doximity Ask and Scribe?
Doximity Ask helps physicians handle tasks like drafting letters and looking up clinical evidence, while Scribe listens during patient visits and turns the conversation into structured notes. Both are built into Doximity’s mobile app.
Tell us about a project or milestone involving AI/ML.
Doximity Ask is our AI assistant. It gives physicians quick, contextual medical answers without leaving their workflow. I lead the iOS AI Pod, and working on this has been one of the fastest-moving projects I’ve been part of. The AI landscape shifts so quickly that features we shipped a few months ago already need rethinking. You learn to build in a way that assumes change. The other thing that shapes how we work is the stakes. This isn’t a chatbot giving restaurant recommendations. If a doctor gets a bad answer, that affects patient care. So we build with that weight, and it makes the team sharper. Doximity Ask fits into a larger strategy of making Doximity the platform doctors use every day, with AI as the thing that keeps them coming back because it actually saves them time and helps them do their job better.
“Doximity Ask fits into a larger strategy of making Doximity the platform doctors use every day, with AI as the thing that keeps them coming back because it actually saves them time and helps them do their job better.”
Doximity’s Approach to AI Development and Ownership
Are there any unique approaches or philosophies to AI or data management that you or your team have?
We recently stood up an agentic development initiative with four pillars: safety, workflow, integration and efficiency. I shepherd the workflow pillar, which focuses on context assembly, development bootstrap and company-shared skills our agents will use. The core idea is that AI tools are only as good as the context you give them. An agent writing code without knowing your patterns, conventions and architecture is just writing generic code. So we’re building systems that give agents layered context: org-level engineering philosophy, team-level patterns and repo-level specifics. We also run an iOS AI council that makes decisions about how our mobile engineers adopt those tools. It’s not top-down mandates. It’s engineers figuring out what actually works and standardizing from there.
What technical aspects of your work are you proud of, either individually or as a team?
I’m proud of being a part of our Doximity Ask project, the AI assistant we built into the Doximity platform. A doctor asks a clinical question and gets a researched, accurate response in seconds. I worked on the iOS side, but getting to watch our back-end team build the full pipeline was eye-opening. Document ingestion, retrieval-augmented generation, model orchestration, retrieval, ranking and generation, all running in sequence so the output feels instant. You don’t appreciate how much is happening behind the scenes until you see it layer by layer.
We’re up against some well-known companies in this space, and we’re keeping pace. Doctors are using it and giving us amazing feedback. It’s helping them take care of their patients. Think of it like a research team behind a reference desk. One person pulls the right documents, another checks relevance, another organizes the findings, and the person at the desk delivers a clean answer. The doctor only sees the desk. But behind it, there’s a full pipeline of retrieval, ranking and generation all coordinating in real time. That’s what the architecture is doing every time someone asks a question.
What It’s Like to Work on AI at Doximity
What do team members working with AI/ML at Doximity get to do that they might not be able to elsewhere? Why should an applicant interested in AI join your team?
Engineers here get real ownership over how AI gets adopted across the company. I run local agentic workflows that take a Jira ticket and use sub-agents and skills to build features through to a pull request. I’m given the freedom to experiment and refine my own workflows. The company stood up an agentic development initiative with dedicated shepherds, a Jira project and weekly syncs to ensure we continue moving in the right direction. You’re not fighting for permission to use AI tools. Leadership is actively asking how to make them better.
For mobile specifically, we’re solving problems most companies haven’t started thinking about, like how agentic coding works when you’re tied to Xcode and simulators. Most of our team runs Claude Code, but the company provides access to ChatGPT, Gemini and Cursor, so each developer can work with the tool and workflow that fits them best. If you want a company that embraces the use of AI, without forcing it upon you, Doximity is a great place to be with a team of strong engineers solving real problems.
Frequently Asked Questions
What does Doximity do?
Doximity provides a digital platform that helps physicians collaborate with colleagues, stay current on medical news and research, manage their careers and on-call schedules, streamline documentation and administrative work, and conduct virtual patient visits.
How does Doximity use AI in its products?
Doximity uses AI in physician-facing products such as Doximity Ask and Scribe. Doximity Ask helps physicians draft letters and look up clinical evidence, while Scribe listens during patient visits and turns conversations into structured notes. AI is also used across product development workflows, including tools and custom skills that support tasks like ticket creation and knowledge sharing.
How does Doximity ensure AI is safe for healthcare use?
Doximity says the stakes are high because incorrect answers can affect patient care, so teams build with that responsibility in mind. The company also describes an agentic development initiative built around four pillars — safety, workflow, integration and efficiency — and says its approach emphasizes giving AI tools layered context, including organization-level philosophy, team patterns and repository-specific details, rather than letting them operate generically.
What software tools do technologists use at Doximity?
Technologists use commercial tools like Claude Code, Cursor, ChatGPT and Gemini. The iOS team also built a custom skill to automate Jira ticket creation using Figma and GraphQL schemas.
