Senior Software Platform Engineer, MLOps

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. We are looking for a Senior Software Platform Engineer who works at the intersection of distributed systems and MLOps. You will build and run key parts of our AI and data infrastructure, which our customers build their science on and which every other engineering team builds against. You will work closely with Applied AI engineers, data engineers, and platform teams to deliver the cloud services and MLOps infrastructure behind production-grade AI/ML workflows. You will own major parts of the AI platform end to end, from design through production. You will work with the Lead Platform Engineer and Applied AI teams on how the pieces fit together, and help shape the platform’s direction through design reviews. The work covers the model and prompt lifecycle, evaluation and observability, tenant and data boundaries around prompts and retrieval, and the cost and latency controls that keep production AI economical at scale. This is a high-impact seat for an engineer who has shipped AI/ML infrastructure that real external users depend on, not only internal tooling. The model serving, MCP, and agent capabilities you build are used directly by scientists at the world's largest pharmaceutical companies, running against their own data under their compliance obligations. If that scope appeals to you, and you want to turn ambitious scalability and cost targets into working systems inside a regulated environment, we'd love to talk to you. What You Will Do - Design, build, and operate the services and APIs that let customers run models and agents against their own scientific data, and that Applied AI and data engineering teams build against internally. - Own the model and prompt lifecycle in your area across Databricks MLflow and AWS Bedrock, including registration, versioning, asset bundles, staged promotion, rollback, and multi-model serving. - Build the inference path for real-time and batch AI workloads, including routing, batching, caching, concurrency control, capacity planning, handling of large binary inputs such as instrument images, and graceful degradation under load. - Integrate LLMs into production systems using retrieval-augmented generation (RAG), and build agent capabilities such as tool and function calling, MCP-based tooling, and agent runtimes. - Build security into the platform with our security team, including guardrails, prompt-injection and tool-abuse defenses, PII and PHI handling, and hard tenant data boundaries across prompts, retrieval, and tool calls. - Build the evaluation and quality tooling that makes AI shippable: offline and online eval harnesses, golden datasets, regression gates in CI, A/B and shadow deployment, and drift and hallucination detection in production. - Set up monitoring, alerting, logging, and distributed tracing for the services you own, and define SLIs and SLOs that make sense for systems whose outputs are probabilistic. - Build for the reproducibility and lineage a validated pharma environment requires: versioned data, code, prompts, and model artifacts, with an audit trail that holds up to customer and regulatory review. - Write infrastructure-as-code and deployment automation (CloudFormation, AWS CDK) for the services you own, working with the team that owns production deployments to support multi-tenant infrastructure, online upgrades, and on-demand compute. - Own production readiness for your services with Applied AI engineers, data engineers, and platform teams: performance, reliability, and cost of models in production, plus incident response and runbooks. - Take part in design reviews, write the design docs and technical documentation for your components, and mentor mid-level engineers. - Evaluate new frameworks, serving runtimes, and model