1 day ago
Base Salary
$172k - $258k/yr
Responsibilities
- Develop and own ML systems from data sourcing, cleaning, and labeling through feature engineering, model development, deployment, and monitoring.
- Set ML architecture for model design, serving, and surrounding systems while optimizing accuracy, latency, and throughput for imbalanced threat-detection data.
- Define production model-serving direction using AWS SageMaker, NVIDIA Triton Inference Server, ensemble/KServe patterns, hardened container images, and enrichment and gateway integrations.
- Benchmark alternatives, prototype major decisions, and provide defensible technical recommendations to leadership.
- Establish reproducible ML standards, including versioned datasets, region-partitioned data, and shared experimentation workflows.
- Implement model observability and efficacy measurement using distributed tracing, threshold-independent metrics, raw-payload capture, and regression monitoring.
- Own capacity planning and rollout strategy, including GPU and core throughput, utilization headroom, peak-load provisioning, and phased regional canary or shadow deployments.
- Diagnose production incidents, address structural gaps, review ML code, and mentor engineers.
- Partner with Product, platform engineering, and adjacent teams to shape roadmaps and communicate technical and business implications.
Requirements
- Broad experience with transformer architectures, RNNs, CNNs, generalized linear models, and gradient-boosted trees, with judgment to select appropriate approaches.
- Deep Python proficiency and strong experience with PyTorch, Hugging Face transformers, and NLP tooling; working knowledge of ONNX Runtime, FP16 quantization, and inference acceleration.
- Experience with dense and lexical retrieval, embeddings, vector indexes, approximate-nearest-neighbor search, BM25, TF-IDF, and hybrid retrieval.
- Experience working with datasets exceeding two million examples and highly imbalanced data, including precision/recall trade-offs, threshold selection, and test-set leakage prevention.
- Track record owning production ML systems on AWS, including SageMaker, S3, Athena, Lambda, Glue, Kinesis, and Bedrock, with Terraform, IAM, containers, and Kubernetes deployment.
- Working knowledge of TorchServe, FastAPI, NVIDIA Triton Inference Server, and KServe, including serving trade-offs involving throughput, GPU efficiency, flexibility, and iteration speed.
- Hands-on production experience with CUDA workloads, GPU driver/toolkit/runtime alignment, GPU passthrough in containers, GPU failure debugging, and utilization improvements.
- Fluency with AI-native development tools and LLM application patterns, including OpenAI-style chat-completion and structured tool/function-calling APIs, MCP, and agent frameworks.
- Demonstrated individual-contributor technical leadership, including setting direction, mentoring engineers, and communicating decisions to technical and executive audiences.
- Understanding of sensitive-data handling under Master Service Agreements and compliance requirements.
- Ph.D. or Master's degree in a quantitative discipline such as computer science, statistics, or mathematics, with substantial production ML experience; alternatively, a Bachelor's degree with equivalent depth demonstrated through a longer track record.
Benefits
- Employees are expected to work from the office at least two days per week.
- The position includes benefits and may be eligible for incentive plans in accordance with company policy and local regulations.
- Employment is subject to successful completion of applicable background checks.
Categories
About Mimecast
Mimecast builds cloud services that secure email, web, and collaboration, plus archiving, eDiscovery, and awareness training for organizations. It sells subscription-based cybersecurity and compliance products with deep Microsoft 365 support, used to protect communications and manage human risk. Founded in 2003 and headquartered in London, Mimecast serves over 42,000 customers in 175+ countries and is owned by private equity firm Permira.