Staff Software Engineer, AI Serving (AI Infrastructure)
Role details
Job location
Tech stack
Job description
HubSpot is on the forefront of empowering businesses with cutting-edge AI, and our AI Infrastructure group is at the heart of this initiative. We are seeking a Staff level Software Engineer with a passion for infrastructure and developer experience to join our AI Serving team. The AI Serving team is part of AI Infrastructure and its goal is to own and operate HubSpot's model inference and serving infrastructure end-to-end, ensuring AI models are reliably deployed, scaled, observed, and cost-efficient across all hublets and environments, while also serving as the home for modern Python at HubSpot.
The team owns the serving frameworks and libraries that power synchronous, asynchronous, and offline inference, along with model deployment pipelines, autoscaling, and GPU/CPU resource management. The team's roadmap focuses on improving the reliability and efficiency of batch inference, strengthening model observability, and scaling AI infrastructure cost-effectively. The team also serves as the home for core Python libraries and tooling across HubSpot.
As a Staff Software Engineer on the AI Serving team you will:
- Write and ship production code that directly powers AI capabilities across HubSpot, a crucial building block for user and business impact at scale
- Lead the evolution of our model serving and inference infrastructure, improving reliability, scalability, and cost-efficiency
- Set technical direction for how HubSpot deploys and operates ML models at scale, guiding decisions on architecture, cost optimization, and developer experience
- Work closely with Machine Learning Engineers, Product Software Engineers, and other Infrastructure teams to streamline model deployment and serving
- Gain deep exposure to a wide range of HubSpot infrastructure systems through building and maintaining core Python libraries, an excellent opportunity to learn how HubSpot operates under the hood
- Collaborate with HubSpotters all over the company, whether in engineering, product, design, research, marketing, sales, etc, gaining exposure to how HubSpot works as a business
- Be empowered to show up as your authentic self and choose a work style and location where you can do your best work, whether that's at home, in an office, or a blend of the two, HubSpot may use AI to help screen or assess candidates, but all hiring decisions are always human. More information can be found here. By submitting your application, you agree that HubSpot may collect your personal data for recruiting, global organization planning, and related purposes. Refer to HubSpot's Recruiting Privacy Notice for details on data processing and your rights.
Requirements
We're not just looking for engineers with many years of experience. Instead we're looking for engineers who also have a track record of unique excellence, leadership, and leveling up those around them.
We work with Python and Java primarily, but experience with specific technologies is secondary to strong programming fundamentals. Check out our engineering blog to learn more.
What we'd like to see:
- A proven track record of directly contributing to and leading infrastructure projects at scale, ideally in serving, deployment, or platform engineering
- The ability to work collaboratively with the team to solve problems, transfer knowledge, and develop overall platform architecture
- The ability to work effectively across teams, driving technical initiatives spanning multiple groups within AI Infrastructure and beyond
- Comfort with ambiguity: you can take on broad, cross-team problems and drive them to resolution without waiting for every detail to be specified
- Coaching and mentorship experience
- A track record in partnering successfully with recruiting to build remarkable engineering teams
- Bonus: Python expertise. Experience with Python tooling, libraries, and developer experience is a strong plus, given the team's role in stewarding Python across AI workloads at HubSpot