AI Engineer - Prompt & Context Engineering (Graph + LLM Comparison) in London

We are hiring an AI Engineer to design and operationalize LLM-driven comparison workflows that produce decision-ready comparison reports across Salesforce orgs and codebases. The role focuses on prompt engineering, context engineering, and agentic orchestration to generate structured outputs (including AST or AST-like representations for code). You will blend deterministic Graph DB reports with non-deterministic LLM insights, and enable human-in-the-loop validation.

Hybrid Working

The places that you work from day to day will vary according to your role, your needs, and those of the business; it will be a blend of Company offices, client sites, and your home; noting that you will be unable to work at home 100% of the time.

Your Role

• Design, iterate, and maintain prompt + context strategies to generate consistent, structured comparison outputs (JSON schemas, diff formats, summaries).
• Build and run agent workflows to orchestrate multi-step comparison tasks (planning, tool calls, retries, fallbacks, state management).
• Produce code-aware comparisons using AST or similar structural representations (e.g., class/method-level diffs, dependency summaries, refactor suggestions).
• Integrate deterministic Graph DB findings (commonality/variation, dependency paths, lineage) into LLM context to improve reliability and traceability.
• Implement output guardrails: schema validation, format constraints, confidence scoring, and consistency checks across runs.
• Support MCP-based integrations where required to connect tools, enforce controlled context, and standardize tool interfaces for agents.
• Partner with Salesforce engineers/architects to validate findings, refine templates, and ensure outputs are reviewable and audit-friendly.

Your Skills

• Strong prompt engineering and context engineering (structured prompting, tool/function calling, grounding strategies, hallucination mitigation).
• Hands-on experience building LLM agents and orchestration (multi-step workflows, state/memory, tool routing, error handling).
• Ability to generate structured code analysis outputs (AST/parse-tree approaches, semantic diffs, dependency extraction, refactoring insights).
• Experience combining deterministic outputs (Graph DB queries, rule checks) with non-deterministic LLM reasoning in a controlled pipeline.
• Python is a strong plus (automation scripts, parsers, report generators, lightweight services).
• Salesforce metadata understanding is a plus (objects/fields, flows, profiles/permsets, dependency concepts, cross-org comparison needs).
• Proven experience with Claude for structured outputs and agentic workflows (prompt patterns, tool-use/function calling, long-context strategies).
• Familiarity with Claude-friendly prompt hygiene: explicit schemas, deterministic formatting, step constraints, and robust self-check prompts.
• Experience using Claude in MCP-based tool ecosystems (defining tools, controlling context sources, safe orchestration patterns).

EEO Statement

We are a Disability Confident Employer (Level 2) under the UK Government’s Disability Confident scheme. As part of our commitment to inclusive recruitment, we will offer an interview to all candidates who declare they have a disability and meet the minimum essential criteria for the role. Please opt in during the application process.