AI Engineer

Own features end to end - data ingestion, retrieval, orchestration, API, and enough Front End to put a usable interface in front of a user.

• Build the evaluation and guardrails alongside the feature: test sets, regression checks, human-in-the-loop review, and a clear answer to "why did the system produce this?"

• Work directly with deal teams, portfolio operations and functional experts. Translate a vague, expert-language problem into something that can be scoped, built and measured.

• Make deliberate build decisions on models and deployment - hosted vs. self-hosted, when fine-tuning is worth it, where data can and cannot go, and what the cost and latency envelope is.

• Raise the engineering bar in the pod: code review, mentoring, sensible standards, keeping the platform’s shared components genuinely reusable.

Substantial software engineering track record (typically 8+ years) as a hands-on senior individual contributor - someone whose seniority shows up in judgement, not distance from the code.

• Production experience with LLM systems, not prototypes: RAG and retrieval design, embeddings and vector search, agent or multi-step orchestration, prompt and context engineering, evaluation.

• Strong core engineering in at least one Server Side language (Python, Java, C#, Go or TypeScript) plus the ability to build the surrounding system - APIs, pipelines, data models, deployment.

• Solid data engineering instincts: SQL, working with messy real-world data, pipeline design, query and schema optimisation where performance matters.

• Enough Front End capability to ship a working interface (React, Angular, Svelte, Vue, Streamlit or similar) - you don’t need to be a design specialist, but you shouldn’t be blocked without one.

• Experience delivering in a regulated, confidential or otherwise trust-sensitive environment, where auditability, access control and data handling are constraints from day one.

• Demonstrated ability to work directly with non-technical domain experts and turn their language into a shipped feature. “, “industry”: “IT”, “baseSalary”: {“@type”: “MonetaryAmount”, “currency”: “GBP”, “value”: {“@type”: “QuantitativeValue”, “maxValue”: “500”, “minValue”: “500”, “unitText”: “DAY”}}, “identifier”: {“@type”: “PropertyValue”, “name”: “[[BRANDNAME]]”, “value”: “[[IDENTIFIER]]”}, “datePosted”: “2026-08-06T13:55+00:00”, “validThrough”: “2026-08-20T13:55+00:00”, “hiringOrganization”: {“@type”: “Organization”, “name”: “Elliott Allen Consulting”, “logo”: “”}, “jobLocation”: {“@type”: “Place”, “address”: {“@type”: “PostalAddress”, “addressRegion”: “England”, “addressLocality”: “London”, “addressCountry”: “United Kingdom”}, “geo”: {“@type”: “GeoCoordinates”, “latitude”: “51.512”, “longitude”: “-0.065”}}, “url”

We are building an in-house AI platform and shipping AI-first features into the way our investment and portfolio teams actually work - diligence, deal screening, portfolio monitoring, research and internal knowledge. Delivery runs through small.AI pods: a handful of senior engineers Embedded close to the business, owning a problem end to end from problem framing through to a production feature people use daily. There is no throwing work over a wall to a separate build team, and no long architecture phase before anyone writes code.

The Job

• Design and build AI-first product features on our internal platform: retrieval over large unstructured document estates, agentic and multi-step workflows, natural-language query over structured data, classification and extraction pipelines.