Artificial Intelligence Engineer

AI Engineer

Contract · On-site (Southern California) · 30 Days Engagement· Immediate Start

ABOUT THE ROLE

Conquer AI is looking for two Senior AI Engineers to embed within a major US financial institution on a fast-paced 30-day proof of concept. This is a hands-on engineering role. You will be on-site at the client's Southern California headquarters for the first two weeks, working in a war room environment alongside their operations and technology teams.

You will pair directly with business users and internal stakeholders, building in real time rather than writing specifications. The engagement is scoped around two high-impact use cases in loan operations, with an aggressive near-term board deadline.

This is not a consulting or advisory role. We need engineers who write code, ship working software, and leave clients self-sufficient.

WHAT YOU'LL BUILD An extraction pipeline that ingests loan documents (financial statements, tax returns, income statements) across multiple formats. Extract key legal terms and financial metrics. Normalise and map to loan-type-specific templates with human-in-the-loop review, confidence scoring, and source citation so reviewers can click through to verify each field against the original document. Target: significantly reduce manual review time per document, with full audit trail.

Standardise financial spreading across CRE, corporate, and speciality lending structures. Build basic ML scoring and weighted financial spread models. Integrate with a legacy core banking system via Excel. Requirements are captured through pair programming, live with the business user, not through handoff documents. Starting with plain vanilla loan structures, building toward more complex arrangements.

RESPONSIBILITIES

• Pair with loan operations SMEs to capture requirements live and build iteratively against real documents
• Design and build agentic AI pipelines for document extraction, normalisation, and structured data output
• Develop financial spreading models with Excel integration for downstream legacy system ingestion
• Implement human-in-the-loop review workflows, feedback loops, and source citation mechanisms
• Train business users through doing: prompt engineering alongside them, not in a classroom
• Deliver train-the-trainer handoff including prompt playbooks, methodology documentation, and champion certification
• Produce board-ready before/after metrics at the end of the engagement

REQUIREMENTS

Must Have

• 5+ years of software engineering experience, with at least 2 years working on production AI/ML systems
• Hands-on experience building agentic AI systems using LLMs (OpenAI, Anthropic, or equivalent)
• Strong Python engineering: you write clean, production-grade code
• Experience with document extraction pipelines including OCR, NLP, and structured data extraction from unstructured sources
• Comfortable working directly with non-technical business users and translating their workflows into working software
• Experience deploying into enterprise cloud environments (Azure preferred) with security constraints
• Strong understanding of prompt engineering, RAG patterns, and LLM evaluation

Nice to Have

• Experience in financial services: banking, lending, asset management, or similar regulated environments
• Familiarity with financial spreading, loan operations, or credit analysis workflows
• Experience fine-tuning or working with small language models for domain-specific extraction tasks
• Exposure to legacy system integration via Excel or similar data bridge patterns
• Prior on-site client-facing delivery experience

THE ENGAGEMENT

Weeks 1-2: On-site at client headquarters in Southern California. War room environment. Building live with the client team.

Weeks 3-4: Continued delivery (remote or on-site as agreed), end-to-end testing, train-the-trainer handoff, board presentation materials.

This is an immediate start engagement. Security clearance, VPN configuration, and sandbox environment access will be completed in Week 1.

There is a clear path to extended engagement (60 and 90-day scope) for the right engineers based on POC outcomes.

ABOUT CONQUER AI

Conquer AI is a specialist agentic AI consultancy working with listed businesses across financial services, asset management, and beyond. We are a team of ~30, Series A funded, with delivery teams in London and the US. Our differentiator is simple: we have been building agentic AI in production, not just talking about it, for years.