octoenergy - Octopus Energy Group

London (GB) - Hybrid3w ago

In Office EMEA Asset Management Fintech Systems Engineer Staff Scientist REST Team Management Reporting SQL Python

Requirements

• Strong systems integration experience - connecting business systems, SaaS tools and data sources via REST APIs, connectors and authentication flows. • A track record of building and shipping data products or internal applications that people actually use, not just pipelines that feed someone else's reports. • Solid Python and SQL, used daily for building, automation and data modelling. • Hands-on experience with a cloud data platform in production (Databricks preferred): workspaces, catalogs, permissions and compute. • Practical experience integrating AI into real tools - LLM APIs, retrieval, agents - with a clear-eyed sense of what works and what's still a demo. • Practical identity and access experience - SSO, service accounts, credential management and access lifecycle - enough to make sensible, secure choices without a security team holding your hand. • Experience with Microsoft 365 / SharePoint integration, or a comparable enterprise document and collaboration stack. • Real experience keeping cloud or AI usage costs under control: you know what drives the bill and how to attribute and cap it. • A pragmatist's instinct for adopting over building where infrastructure is concerned - you'd rather configure something that exists than write something new - paired with a bias toward shipping when it comes to solutions. • Clear communication. You can explain a trade-off to a fund manager and hold your own in a technical review with a central engineering team. • Comfort with real ownership in a small team, where you set the working patterns rather than inherit them. • Full-stack or application development experience, particularly with Streamlit, FastAPI or HTMX. • Familiarity with the Databricks ecosystem - Unity Catalog, Databricks Apps, Lakebase, Genie - or equivalent tooling. • Experience with vector search, embeddings, document AI or retrieval-augmented generation in production. • Experience with Terraform, GitHub Actions or similar, to keep configuration version-controlled and repeatable. • FinOps experience: usage monitoring, budgets, showback and chargeback. • Having supported analysts and non-specialist builders using a shared data environment. • Experience in a regulated or audited setting such as financial services, infrastructure investment or energy. • Familiarity with renewable energy, infrastructure investment or financial services data.

Responsibilities

• Integrate the systems our business runs on - third-party providers, internal tools, our own SharePoint and Microsoft 365 tenant - into our existing Databricks environment, via APIs and existing connectors. • Build data products on top of what you connect: well-modelled, documented, reusable datasets that the business consumes directly through Excel, dashboards and natural-language querying, rather than one-off extracts. • Design and ship internal applications and AI-enabled tools - using Python, Streamlit, Databricks Apps (HTMX), Lakebase, FastAPI or similar - that put data and AI directly into the hands of fund management, asset management and finance teams. • Prototype, evaluate and productionise new AI capability, from document AI and retrieval to agentic workflows, and turn the ideas that prove out into supported products rather than abandoned experiments. • Take an adopt-first approach: work with Octopus Energy Group's data platform and AI teams to reuse capability that already exists, and configure it for OEGEN rather than building parallel versions of it. • Set up practical guardrails inside Databricks and Unity Catalog - schema and workspace access, permissions, environment hygiene - so more teams can use it without stepping on each other. • Bring cost management from reactive to proactive: tagging, budgets, alerting and usage monitoring, so Databricks and AI token spend can be attributed to the right team, project or fund before the invoice arrives. • Get our AI usage properly credentialed and logged - managed keys and access rather than API keys sitting in individual password managers - and provide clean patterns for embedding AI into the tools we build. • Standardise how we build and ship internal apps: shared templates, authentication, deployment and documentation, so a tool remains supportable by someone other than the person who wrote it. • Consolidate the ad-hoc scripts, spreadsheets and no-code workflows we've accumulated into fewer, better-understood integrations that don't need babysitting, and automate the repetitive parts of our own workflow. • Work directly with non-technical teams to understand their problems and translate them into things we can build.