AI Engineer, Citi - Job Offer

Citi

Location:
United Kingdom, London

Category:
IT - Software Development

Contract Type:
Employment contract

Salary:

Not provided

Save Job

Apply Position

Job Description:

By joining Citi London, you will not only be part of a business casual workplace with a hybrid working model (up to 2 days working at home per week), but also receive a competitive base salary (which is annually reviewed), and enjoy a whole host of additional benefits.

Job Responsibility:

Productise GenAI: Deliver LLM-powered workflows for internal users and clients using hosted or private models
Build Agentic Systems: Design autonomous agents with reasoning, planning, and tool-use capabilities
implement multi-agent coordination for complex workflows
embed guardrails and deterministic fallbacks
Architect advanced RAG pipelines: Deploy hybrid retrieval strategies, incorporating semantic search, query routing, contextual compression, and re-ranking
optimize for accuracy, latency, and cost
ensure auditability with proper citations and provenance tracking
Establish evaluation frameworks: Define comprehensive assessment methodologies covering contextual relevance, hallucination detection, and grounding. Produce model-risk documentation aligned to Citi standards
Set technical standards: Create reusable patterns for prompt engineering, orchestration, and model selection
establish best practices for GenAI-powered application development
influence architecture decisions across Innovation Labs
Lead stakeholder collaborations: Translate business requirements into technical roadmaps. Effectively communicate with senior stakeholders and cross-functional partners
Provide technical mentorship: Provide technical guidance to data scientists and product engineers on model selection, evaluation design and production best-practices
elevate team capabilities through code/architecture reviews and knowledge-sharing
Continually evolve: Stay at the forefront of AI advancements and industry best practices, driving a culture of continuous improvement and innovation within our systems

Requirements:

Applied Machine Learning Acumen: Strength in model capabilities/limits, prompting patterns, function/tool-calling, embeddings selection, and vector semantics (ANN indexes, HNSW/IVF, hybrid search, rerankers)
Agentic AI Design: Proven experience building multi-tool, multi-step agents with deterministic fallbacks, safety checkpoints, and recovery strategies
Advanced RAG Expertise: Hands-on expertise with semantic chunking, query decomposition, hybrid search, metadata filtering, and reranking
understanding of retrieval optimization and result synthesis strategies
Evaluation methodologies: Experience designing robust evaluation frameworks (both manual and automated) including golden sets, reference-free evals, benchmark testing, and online metrics
Software Engineering Toolkit: A commitment to best practices, utilizing design patterns to architect scalable, maintainable solutions, and a history of leading impactful code reviews that elevate team quality
Expert Python Craftmanship: Deep expertise in Python, evidenced by building high-performance applications, developing bespoke modules, and seamlessly integrating Python with diverse systems using advanced features and libraries
Stakeholder Communication: Ability to articulate complex technical concepts to both technical and non-technical audiences
comfortable presenting to senior leadership
skilled at translating business needs into technical solutions

Nice to have:

Experience working with multimodal data (text, images, and documents)
Fine-tuning knowledge (LoRA, QLoRA, PEFT) and understanding of when to fine-tune versus retrieve or prompt-engineer
Document processing expertise for PDFs, tables, and structured data extraction
Knowledge graphs or entity store for enhanced grounding and retrieval
Structured tool integration (SQLs/APIs) for complex task execution
Familiarity with AI governance and policy enforcement mechanisms
Proficiency leveraging coding agents (GitHub Copilot, Claude Code, Devin) to accelerate development velocity and code quality
Background in financial services industry beneficial, but not required

What we offer:

27 days annual leave (plus bank holidays)
A discretional annual performance related bonus
Private Medical Care & Life Insurance
Employee Assistance Program
Pension Plan
Paid Parental Leave
Special discounts for employees, family, and friends

Additional Information:

Job Posted:
October 11, 2025

Employment Type:

Fulltime

Work Type:

On-site work

View All Jobs In This Company

Job Link Share:

AI Engineer