Director, AI Architect

Director, AI Architect

About Headspace

Headspace exists to provide every person access to lifelong mental health support. We combine evidence-based content, clinical care, and innovative technology to help millions of members around the world get support that's effective, personalized, and truly accessible whenever and wherever they need it.

At Headspace, our values aren't just what we believe, they're how we work, grow, and make an impact together. We live them daily: Make the Mission Matter, Iterate to Great, Own the Outcome, and Connect with Courage. These values shape our decisions, guide our collaborations, and define our culture.

About the Role

Our mission is to transform mental healthcare from an episodic need to an everyday care practice by offering an AI-powered, clinically safe, and personalized Mental Health Companion. AI is central to how we fulfill that mission at scale.

As our Director, AI Architect, you will own the technical vision and execution of Headspace's AI-first service transformation, designing the systems, patterns, and practices that embed intelligence across every layer of our product and platform.

Reporting directly to the Chief Product & Engineering Officer, you will operate at the intersection of applied research, ML science, and platform engineering. This is a rare opportunity to shape how a category-defining mental health company thinks about, builds with, and deploys AI, responsibly, scalably, and in ways that genuinely improve member outcomes.

Responsibilities

• Define and lead Headspace's overarching AI architecture strategy, establishing the foundational patterns, platforms, and principles for an AI-first service transformation across our portfolio of streaming content, conversational AI, coaching, therapy and psychiatry services.
• Architect end-to-end AI systems: including LLM-powered features, agentic workflows, retrieval-augmented generation (RAG), and real-time personalization, with a relentless focus on reliability, safety, and member impact.
• Partner directly with the executive team, product leadership, and senior engineering stakeholders to align AI strategy with company goals, translating business opportunities into concrete technical roadmaps.
• Author company-wide technical specs that establish AI design principles, evaluation frameworks, guardrails, and reusable platform components.
• Drive responsible AI practices across the organization, including model evaluation, bias mitigation, explainability, data governance, and compliance with evolving regulatory standards relevant to health tech.
• Lead the selection, evaluation, and integration of AI/ML infrastructure, including model providers, vector databases, orchestration frameworks, and MLOps tooling, balancing build vs. buy decisions with long-term strategic implications.
• Collaborate with Data Science, ML Engineering, and Product teams to ensure AI systems are grounded in high-quality, privacy-preserving data pipelines and continuously improve through rigorous feedback loops.
• Establish AI engineering standards and best practices across squads, from prompt engineering and context management to model versioning, observability, and production monitoring.
• Mentor and elevate engineers, ML practitioners, and technical leads across the organization, helping teams build confidence and competency in applied AI development.
• Serve as Headspace's internal and external thought leader on AI, representing the company's technical vision in recruiting, partnerships, and the broader industry.
• Identify and evaluate emerging AI capabilities (reasoning models, multimodal systems, fine-tuning approaches) for near-term applicability to Headspace's roadmap.

Required Skills

• 10+ years of software engineering experience, with at least 4 years focused on the design and delivery of production AI/ML systems at scale.
• Deep expertise in modern AI architectures, including LLMs, RAG systems, embedding pipelines, agentic frameworks, and real-time inference, with hands-on experience moving these from prototype to production.
• Proven ability to define AI strategy at an organizational level: translating ambiguous business challenges into technical roadmaps, influencing executive stakeholders, and driving alignment across cross-functional teams.
• Strong command of responsible AI principles: safety, fairness, explainability, data privacy, and the unique ethical considerations of AI in health and wellness contexts.
• Extensive experience with cloud-native AI infrastructure (AWS, GCP, or Azure), containerized deployment (Kubernetes), and MLOps practices including model serving, monitoring, and evaluation pipelines.
• Demonstrated ability to evaluate and integrate third-party AI providers, orchestration frameworks (e.g., LangChain, LlamaIndex, or similar), and vector/embedding database systems.
• Exceptional communication skills: you can articulate complex AI trade-offs clearly to both technical engineers and non-technical executives, and write specs that bring entire organizations along with you.
• Ownership mindset: you are comfortable navigating ambiguity, making consequential architectural decisions with incomplete information, and taking accountability for outcomes across teams.

Preferred Skills

• BS/MS/PhD in Computer Science, Machine Learning, or a related field, or equivalent practical experience.
• Experience in digital health, wellness, or a similarly regulated consumer domain, with familiarity with HIPAA, data minimization practices, and the heightened standard of care for AI in sensitive user contexts.
• Background in fine-tuning, RLHF, or domain-adapted model training for specialized consumer applications.
• Experience with conversational AI, dialogue systems, or AI-powered coaching/companionship products.
• Familiarity with Server-Driven UI (SDUI) and how AI-driven personalization integrates with dynamic, schema-based rendering across web and mobile clients.
• Track record of building AI evaluation frameworks — including automated evals, red-teaming, and human-in-the-loop review pipelines — to maintain quality at scale.
• Experience driving AI governance initiatives, including model cards, audit trails, and cross-functional risk review processes.

Location

This is a hybrid role located in our San Francisco office, working 3 days per week from the office.

Compensation & Benefits

Base Salary: $230,000–$287,500 + equity + benefits

Our salary ranges are based on the job, level, and location, and reflect the lowest to highest geographic markets where we are hiring for this role within the United States. Within this range, individual compensation is determined by a candidate's location as well as a range of factors including but not limited to: unique relevant experience, job-related skills, and education or training. Your recruiter will provide more details on the specific salary range during the hiring process.

At Headspace, base salary is but one component of our Total Rewards package. We're proud of our robust package inclusive of:

• Base salary
• Stock awards
• Comprehensive healthcare coverage
• Monthly wellness stipend
• Retirement savings match
• Lifetime Headspace membership
• Generous parental leave

Additional details about our Total Rewards package will be provided during the recruitment process.

Why You'll Love Working Here

• A mission that matters—with impact you can see and feel
• A culture that's collaborative, inclusive, and grounded in our values
• The chance to shape what mental health care looks like next
• Competitive pay and benefits that support your whole self

Diversity, Equity, Inclusion and Belonging

Headspace is committed to bringing together humans from different backgrounds and perspectives, providing employees with a safe and welcoming work environment free of discrimination and harassment. We strive to create a diverse & inclusive environment where everyone can thrive, feel a sense of belonging, and do impactful work together.

As an equal opportunity employer, we prohibit any unlawful discrimination against a job applicant on the basis of their race, color, religion, gender, gender identity, gender expression, sexual orientation, national origin, family or parental status, disability, age, veteran status, or any other status protected by the laws or regulations in the locations where we operate. We respect the laws enforced by the EEOC and are dedicated to going above and beyond in fostering diversity across our workplace.

Applicants with disabilities may be entitled to reasonable accommodation under the terms of the Americans with Disabilities Act and certain state or local laws. A reasonable accommodation is a change in the way things are normally done which will ensure an equal employment opportunity without imposing undue hardship on Headspace. Please inform our Talent team by filling out this form if you need any assistance completing any forms or to otherwise participate in the application or interview process.

Headspace participates in the E-Verify Program.

Important Information

All member records are protected according to our Privacy Policy. Further, while employees of Headspace (formerly Ginger) cannot access Headspace products/services, they will be offered benefits according to the company's benefit plan. To ensure we are adhering to best practice and ethical guidelines in the field of mental health, we take care to avoid dual relationships. A dual relationship occurs when a mental health care provider has a second, significantly different relationship with their client in addition to the traditional client-therapist relationship—including, for example, a managerial relationship.

As such, Headspace requests that individuals who have received coaching or clinical services at Headspace wait until their care with Headspace is complete before applying for a position. If someone with a Headspace account is hired for a position, please note their account will be deactivated and they will not be able to use Headspace services for the duration of their employment.

Further, if Headspace cannot find a role that fails to resolve an ethical issue associated with a dual relationship, Headspace may need to take steps to ensure ethical obligations are being adhered to, including a delayed start date or a potential leave of absence. Such steps would be taken to protect both the former member, as well as any relevant individuals from their care team, from impairment, risk of exploitation, or harm.

For how we will use the personal information you provide as part of the application process, please see: https://www.headspace.com/applicant-notice