Lead platform engineering for internal shared data and AI capabilities, including ingestion, integration, canonical APIs, and agent workflow platforms with orchestration, monitoring, evaluation, observability, and guardrails. The team builds scalable, reliable platform services for LexisNexis Intellectual Property Solutions, enabling data-driven innovation and production-grade AI workflows across multiple product domains.

Work with TensorFlow or PyTorch, Python, and related computer vision/graphics libraries on neural network architectures (convolutional, recurrent, transformer, diffusion-based). The Visual Computing Group (VCG) researches and develops generative models for game rendering and live video streaming to advance state-of-the-art runtime efficiency, visual quality, and latency.

You will use TensorFlow or PyTorch, Python, and CV/graphics libraries, plus 3D engines like Unreal or Unity, to develop generative models for graphics and video. As part of the Visual Computing Group, you'll design neural networks and ML for PlayStation's game rendering and live

Databricks★3.0 (21)· London

You will work with Python, Scala, JavaScript/TypeScript, Apache Spark, and cloud ecosystems (AWS, Azure, GCP) to build production-grade data and AI solutions on the Databricks platform. In this hands-on role, you lead architecture and design decisions while embedding with customer teams to deliver end-to-end data pipelines, ML/AI models, and user-facing applications.

Design and develop AI agents using AWS AgentCore, Amazon Bedrock, Python, and TypeScript. Build backend services with AWS Lambda, API Gateway, Step Functions, DynamoDB, and IaC tools like Terraform or AWS CDK for enterprise multi-agent workflows and production monitoring.

This role focuses on designing, building, and operating enterprise AI agents using AWS AgentCore, Amazon Bedrock, Python, and TypeScript. You will implement RAG, tool calling, and multi-agent workflows, developing backend services with AWS Lambda, API Gateway, Step Functions, and DynamoDB. The position also involves establishing production AIOps with observability, monitoring, and CI/CD pipelines using Terraform or AWS CDK.

The Head of AI works with Python, NLP pipelines, LLMs, knowledge graphs, and clinical ontologies like SNOMED CT and ICD. This role leads a team building Dyad's Semantic AI platform combining structured knowledge with generative AI to improve healthcare delivery, with products such as BetterLetter for clinical correspondence processing. The position designs and operationalises the graph-integrated generative architecture for production systems handling unstructured clinical text, including evaluation and compliance frameworks.

You will develop AI agents using AWS AgentCore, Amazon Bedrock, Python, and TypeScript. You will build and operate enterprise multi-agent workflows with RAG, tool calling, serverless AWS services, and production AIOps. The role involves designing, deploying, and monitoring scalable AI architectures with CI/CD and Infrastructure as Code.

Build streaming and batch pipelines on Kafka/Redpanda and a time-series store (ideally ClickHouse) with a lakehouse architecture, using Python, SQL, Docker, Terraform, and CI/CD. You will own data contracts, schema evolution, and governance for the Central Data Team building the Keyrock Data Platform that serves trading desks, wealth/asset management, risk, and compliance.

Build and validate simulation infrastructure for large-scale ML systems, modelling compute, memory, and communication behavior for training and inference. Work with Python, C++, or Rust; preferred tools include PyTorch, JAX, TensorFlow, NCCL, XLA, CUDA, and networking technologies like InfiniBand and NVLink. The role involves performance benchmarking, distributed execution across accelerators, and cross-functional partnership with hardware, software, networking, and ML teams.

You will design, train, and deploy deep learning models (PyTorch/TensorFlow) alongside classical computer vision pipelines (OpenCV, geometric vision) into Python-based Docker microservices and ONNX Runtime C++ desktop applications. You will work on Catapult's centralized Data Science team to build production-ready algorithmic features for sports-performance analytics used by elite professional teams.

LexisNexis Legal & Professional★3.6 (1k)· London

You will lead platform engineering teams delivering shared data and agentic platform capabilities (ingestion, integration, canonical APIs, agent workflows, orchestration) with production-grade reliability, observability, security, and AI guardrails at LexisNexis Legal & Professional. You oversee the organization building and operating scalable internal platforms—data services and AI-workflow infrastructure—consumed by multiple product domains. The role manages engineering teams of unspecified size.

RELX Group★3.8 (133)· London

Lead engineering for shared data platform and agentic platform capabilities, including ingestion, integration, metadata, lineage, APIs, security controls, and AI workflow orchestration with production-grade monitoring. The team builds and operates platform services that enable trusted data and AI-driven workflows for LexisNexis Intellectual Property Solutions.

Design and build fault-tolerant, high-performance systems for serving LLMs, working with Python, Go, Rust, or C++, Docker, Kubernetes, AWS, or GCP, and LLM serving frameworks.

Catapult Sports★3.8 (4)· London

Work with Python, C++, PyTorch/TensorFlow, OpenCV, Docker, AWS, ONNX Runtime, and TensorRT to develop computer vision algorithms for sports performance analytics. As part of Catapult’s Data Science team, you’ll build end-to-end pipelines from video ingestion to production inference, delivering insights for professional teams.

You'll work with Python, Rust, ONNX, Chalk.ai, and GCP (Kubernetes, BigQuery, Agent Platform Endpoints). The team productionizes 10+ ML models for route assignment, delivery prediction, and

You will build and own infrastructure using Linux, AWS, Terraform, CI/CD, Python (Go, Rust, C++, Java beneficial), PostgreSQL, Redis, Kafka, and ClickHouse. Your team builds real-time intelligence systems for global financial markets, combining ML, HPC, and systematic trading. This is a founding role with direct impact on live trading infrastructure.

You will work with Python, modern ML frameworks, and large-scale training workflows. As a Senior Research Engineer on the Code World Models team, you will design pre-training and mid-training experiments for code-centric foundation models, building data pipelines and evaluations to improve how models understand software behavior and developer tools.

You will work with AWS, Databricks, and tools for RAG pipelines, LLM orchestration, vector search, and evaluation workflows. This role focuses on building and owning the cloud-native platform infrastructure for production-grade conversational and generative AI products.

Build and scale distributed training systems using Megatron, DeepSpeed, and NCCL, optimizing throughput and GPU utilization for large model pre-training. You will work closely with researchers to operate large-scale training runs across thousands of GPUs for foundation models at Reflection, an open-intelligence research lab.

Fidelity Investments★4.0 (5.1k)· London

Use Rust, Python, SQL/Snowflake, and blockchain data tools (The Graph, Dune) to build research infrastructure and develop yield-generating DeFi strategies. You will extend backtesting systems and implement on-chain analytics for the Quantitative Strategies team within Fidelity Digital Assets™.

Rivian★2.7 (596)· London

Design and implement learning-based prediction and planning systems using Python, C++, PyTorch, and TensorRT/ONNX for production autonomous driving. Develop hybrid planning architectures combining transformer-based models with classical methods like MPC and graph-based planning. This role is on Rivian’s Autonomy team in London, delivering end-to-end from data and model development to in-vehicle validation.

You will work with JAX, PyTorch, TensorFlow, ADK, MCP, agentic frameworks, LLM ecosystems, and GCP. Isomorphic Labs applies frontier AI to drug discovery, building on AlphaFold to accelerate rational drug design with predictive and generative AI models. You will secure the AI-first platform and autonomous agentic workflows that power the drug discovery pipelines.

Work with Python, Azure (Azure AI Foundry, APIM, Entra ID, RBAC), CI/CD, observability tooling and LLMOps to design and build the Agentic AI Platform — shared standards, templates and reference implementations that enable application teams to safely deploy and operate AI agents at enterprise scale.

Genius Sports★4.4 (37)· London

You will use Python (or R, Matlab) and machine learning algorithms such as neural networks and Monte Carlo simulation, with MLOps via Gitlab, AWS, and Docker. Your multidisciplinary Agile team develops predictive models that calculate sports outcome probabilities for Genius Sports’ live data products.