Ci&T
Digital Transformation Consulting
See how your profile stacks up against this role.
We compared the job requirements to your profile to show where you're strong and where you fall short.
At CI&T, we help large enterprises transform the potential of AI into real business impact with AI Deployment, AI-native execution, and tech-integrated business solutions.
With 30 years of experience in technological transformation, we accelerate innovation with expertise in Agentic SDLC, Application modernization, Data & AI, Martech and Business strategy.
We are 8,000 CI&Ters across more than 25 countries, collaborating to build solutions with real impact. AI is already part of how we work, evolve, and innovate every day.
We are seeking an AI Engineer to lead the design and development of agentic AI solutions that map, automate, and optimize mission-critical business processes in the mortgage servicing domain. This role focuses on escrow management at Top 5 Independent Mortgage Company in the US —a complex, highly regulated process with significant compliance and operational impact.
You will partner closely with a Business Analyst and Technical Subject Matter Expert to translate domain knowledge into AI-powered workflows. Using Claude and Claude Code, you will prototype and build agents that understand escrow operations deeply enough to handle edge cases, regulatory constraints, and process variations. This is not a generic AI role; it requires genuine curiosity about how mortgage servicing works and the discipline to bridge human knowledge and machine learning.
Responsibilities
Domain Intelligence & Process Documentation:
- Partner with the BA and SME to understand escrow management workflows across setup, monthly collection, tax/insurance disbursement, annual analysis, and exception handling. Create executable documentation (process maps, decision trees, regulatory checkpoints) that forms the foundation for agent development.
Agentic System Design & Prototyping:
- Design multi-step agentic workflows that handle domain-specific reasoning—escrow analysis logic, RESPA/Reg X timing requirements, disbursement sequencing, and exception resolution. Use Claude's capabilities (extended thinking, tool use, multi-turn reasoning) to solve problems that require both rule-based logic and contextual judgment.
Implementation & Validation:
- Implement agents and supporting infrastructure using Claude Code, building test harnesses that validate correctness against domain rules and regulatory constraints. Work with compliance and operations teams to field-test agents on real scenarios.
Knowledge Synthesis & Handoff:
- Package domain insights, decision logic, and agent prompts in reusable form so future projects (other mortgage processes, other clients) can build on this foundation. Create runbooks for domain-specific agent tuning and error diagnosis.
Technical Foundation:
- +5 software development experience (Python, TypeScript, or Go preferred). Comfort with API integration, data models, and system design. Hands-on experience with LLM-based applications—prompt engineering, tool integration, multi-turn workflows, and error handling.
AI & Agentic Systems:
- Demonstrated experience building or deploying agentic systems (using frameworks like LangGraph, Semantic Kernel, or custom orchestration). Ability to design control flows for multi-step reasoning, handle uncertainty, and design appropriate guardrails. Strong grasp of prompt design, retrieval-augmented generation (RAG), and LLM limitations.
Domain Translation & Communication:
- Ability to extract domain logic from human experts (SMEs, BAs) without requiring formal specifications. Demonstrated success translating business rules into testable, executable requirements. Comfort asking clarifying questions and identifying edge cases in workflow descriptions.
Regulatory & Risk Mindset:
- Experience in regulated industries (mortgage, fintech, healthcare) or demonstrated understanding of compliance requirements, SLAs, and audit trails. Ability to recognize when an AI-driven approach is unsuitable and advocate for rule-based logic instead.
Nice to Have
After you apply, unlock the direct contact details of the people who actually make the call. A quick follow-up makes you 5x more likely to land an interview.
Marcus Rivera
Chief Revenue Officer
InTalent Asia
Ci&T
Cloudera
Ci&T
Ci&T
Ci&T
Ci&T
Ci&T
