London, United Kingdom
Internship Explorer
Internship Explorer
This 13-month sandwich placement offers a penultimate-year undergraduate the opportunity to contribute directly to the design and implementation of Engelhart’s central Agentic AI platform – a strategic, firm-wide initiative led by the Tech Data, AI & UX team. The role provides hands-on exposure to the cutting edge of enterprise AI, working alongside experienced data engineers, architects and quant developers in a live, production-grade environment. You will work closely with the AI & Data Engineering team to build core agentic infrastructure, integrate tooling via AWS Lambda and MCP servers, develop agent skills, and configure knowledge bases across user, chat, agent, and application contexts. You will explore and evaluate open-source and proprietary connectors within the rapidly evolving LLM and Generative AI ecosystem, and help roll out individual productivity tools such as Microsoft Copilot and Claude Code. All solutions will be hosted securely within Engelhart’s internal network, with models deployed in a dedicated VPC and governed access to firm data, with a strong preference for AWS-native and Bedrock-based architectures. Please note that to be eligible for consideration, candidates must be enrolled on a sandwich year degree and available to commence their placement in July 2026. This will be a full-time role, owning the following responsibilities: * Assist in the set-up and configuration of core agentic AI infrastructure, including orchestration layers, agent runtimes, and supporting services, deployed securely within Engelhart’s AWS VPC. * Build and maintain agent tooling integrations via AWS Lambda functions and MCP (Model Context Protocol) servers, enabling agents to interact with firm data, APIs, and internal systems. * Develop agent skills and configure knowledge bases, including document stores, vector databases (e.g. Postgres with pgvector, AWS Bedrock Knowledge Bases), and retrieval pipelines for user, chat, agent, and application contexts. * Research, evaluate, and help integrate open-source and proprietary accelerators, connectors, and frameworks (e.g. LangChain, LlamaIndex, n8n, etc) into the platform, assessing fitness for purpose in a regulated, IP-sensitive environment. * Support the deployment and end-user adoption of individual AI productivity tools, including Microsoft Copilot and Claude Code, ensuring secure configuration and appropriate access controls within Engelhart’s internal network. * Contribute to technical documentation, architecture diagrams, and internal knowledge sharing on AI tooling, design decisions, and implementation patterns to support team learning and continuity. * Collaborate with engineers, data scientists, and business stakeholders to iteratively test and improve agentic workflows, gathering feedback and translating it into technical improvements.
