Principal Platform Architect, TetraScience

• 12+ years in software engineering, with at least 5 at staff or principal level in a SaaS platform or data infrastructure context
• Deep architecture ownership in at least one of the two fingerprint profiles above, with meaningful range across the other. Coverage of a majority of the eight domains is the bar
• Demonstrated ownership of enterprise authentication and authorization systems at scale: SAML, OIDC, fine-grained RBAC across a multi-tenant SaaS product. You have been the person who got paged when auth broke, not just the person who designed it
• Hands-on experience with AI/ML serving infrastructure: you have built and operated model inference pipelines under production load
• Search architecture experience: you have designed and operated a search platform that handles diverse query types (keyword, semantic, or hybrid) across large structured or semi-structured datasets
• Hands-on experience with data lake architectures at scale: Delta Lake or Apache Iceberg, schema evolution patterns, partition pruning, and the trade-offs between query performance and storage cost
• Infrastructure fluency on AWS with Kubernetes or ECS. You can read a cost anomaly report, trace it to a root cause, and produce an action within the same week
• Ability to write and defend architecture decisions: RFCs, trade-off documents, design reviews
• Strong cross-team communication. You can write a document that produces alignment without a follow-up meeting to explain the document
• Comfort operating across strategy, architecture, and operations in the same week: setting a multi-year architecture direction and reviewing a runbook gap are both in scope

Desirable

• TetraScience is building the scientific data and AI cloud for biopharma. The platform is the innermost foundation for developing, delivering and operating our enterprise grade, secure, compliant Scientific Data and AI capabilities customers rely on
• In this role, you will own the platform architecture, evolution and growth scaling across Enterprise Platform, Scientific Search, AI/ML Ops, Developer Platform, Developer Productivity, Lakehouse Platform, Partner Integrations and Cloud Infrastructure. This is a senior IC leadership role. You set technical direction, own the decisions that cross team boundaries, and close architectural gaps before they become business risks
• After achieving strong product-market fit and traction, we are entering a growth scaling phase where we are expanding our industry partnerships and developer experience to rapidly build the foundations of AI-native scientific data and workflows in production
• The scope of this role is intentionally broad. We are looking for experienced candidates who cover a majority of these areas. Strong candidates bring deep fingerprints in one of two architectural profiles, with meaningful range across both:
• Enterprise Data & AI Platforms: Multi-tenant architecture, RBAC/ABAC, IAM, tenancy models, observability platforms, internal builder platform, cost governance
• Data, Knowledge, and Developer Products: Search, Semantic layer foundations, knowledge and ontology layers and products, external developer platforms
• Enterprise Platform: Tenancy, IAM, compliance and admin control plane that enterprise customers use to govern their scientific data environment: SSO/SAML/OIDC, fine-grained RBAC, multi-tenant isolation, UI infrastructure, and tenant onboarding
• Scientific Search: Search architecture spanning keyword, semantic, and hybrid retrieval across scientific data, instruments, and metadata: relevance standards, indexing pipeline, and the infrastructure that makes search a reliable product surface
• AI/ML Ops: Model serving, agentic infrastructure primitives, embedding services, and the MLOps standards that keep scientific AI outputs traceable and operable under production load
• Developer Platform: The internal paved road: CI/CD standards, golden path tooling, SDK design principles, and the adoption metrics that prove it works
• Developer Productivity: Developer throughput as a first-class metric: toolchain ownership, local/prod environment parity, and friction reduction from commit to deployment
• Lakehouse Platform: Scientific data lake architecture, schema evolution, IDS design standards, and the data access layer that AI workloads and downstream pipelines depend on
• Partner Integrations: Integration architecture for lab instrument vendors and AI model partners: reference patterns, security boundaries, and the developer experience that enables self-service onboarding
• Cloud Infrastructure: Production architecture, cost governance, and the observability layer from infra signal to customer-visible service health
• What success looks like in year one:
• Authn/Authz architecture is documented, consistent across services, and passing enterprise security reviews without heroics from a single engineer
• AI/ML infrastructure has a clear architecture and roadmap for MLE inference and training use cases, with strong operational telemetry and cost visibility
• The developer platform has clear SDKs and a set of standard templates for scientific use cases to start from, with adoption and delivery by multiple scientific use case teams
• Operational excellence based on a clear O11y architecture rolled out, with every production service having SLOs defined, monitored and managed
• Cost governance with customer chargeback attribution architecture and operationalized with the finance and field teams
• Lakehouse platform architecture and operational buildout as a Data Products Platform with strong DX and operational scaling
• Evolve IDS to open standards based schema and encoding with strongly typed data models and schema-on-write enforcement
• Published reference architecture for each partner class (lab instrument manufacturers and AI models), with one partner successfully onboarded against each without bespoke engineering support