Gen AI Engineering and Scaled AI Transformation

Citigroup Inc.; Mississauga; Hybrid; CAD 150,000 - 190,000; Full time

Job summary

Citibank (Switzerland) AG in Mississauga, Ontario, seeks a senior Gen AI leader to own the LLM strategy and enterprise lifecycle governance for the Source to Pay technology group, shaping hosting, privacy, cost, and performance trade-offs.

You will drive hands-on GenAI work with LangChain, LangGraph, LlamaIndex, and Hugging Face, design robust RAG architectures with vector stores, and guide multi‑team execution toward scalable, secure, and auditable AI platforms.

Job description

## Gen AI Engineering and Scaled AI TransformationApplyremote type: Hybridlocations: Mississauga Ontario Canadatime type: Full timeposted on: Posted Todaytime left to apply: End Date: July 24, 2026 (6 days left to apply)job requisition id: 26979423**Role Focus:** **Generative AI Engineering and Scaled AI Transformation for Source to Pay technology group - Hybrid****1. Large Language Model (LLM) Strategy & Technical Authority*** Acts as a **senior technical authority** on Large Language Models, including both **commercial and open‐source ecosystems** (OpenAI, Gemini, Claude, Llama).* Leads **model selection and deployment strategy**, balancing use‐case fit, data sensitivity, cost efficiency, latency, accuracy, and regulatory constraints.* Guides decisions on **hosted vs. private vs. fine‐tuned models**, ensuring optimal trade‐offs between performance, control, and operational risk.* Establishes **enterprise standards for LLM lifecycle management**, including upgrades, regression validation, and decommissioning.**2. Hands‐On GenAI Application & Agentic System Design*** Demonstrates **hands‐on leadership** in building GenAI applications using **LangChain, LangGraph, LlamaIndex, and Hugging Face**, translating experimentation into production systems.* Architects **agentic and multi‐step workflows**, enabling tool‐use, reasoning chains, state management, and orchestration at enterprise scale.* Sets reusable **reference patterns and accelerators** for GenAI adoption across application teams.* Ensures solutions are built with **enterprise-grade reliability, explainability, and extensibility**.**3. Retrieval Augmented Generation (RAG) & Enterprise Knowledge Enablement*** Designs and delivers **robust RAG architectures** that ground GenAI outputs in trusted, auditable enterprise data.* Leads implementation of **vector databases and embedding strategies** (pgvector, Pinecone, Weaviate, FAISS), aligned with data access and security models.* Applies **advanced retrieval techniques** including hybrid search, re‐ranking, metadata filtering, and context optimization to improve response accuracy and relevance.* Ensures RAG solutions support **data lineage, auditability, and regulatory compliance**.**4. Prompt Engineering, Workflow Optimization & Cost Control*** Establishes **prompt engineering and orchestration standards** to ensure consistency, maintainability, and quality across GenAI solutions.* Optimizes GenAI workflows by actively managing **latency, throughput, token cost, and accuracy trade‐offs** in production environments.* Implements **evaluation and experimentation frameworks** to continuously improve output quality and business value.* Drives disciplined use of caching, batching, fallback models, and token optimization techniques.**5. Machine Learning & Model Enablement Foundations*** Applies strong grounding in **ML/DL fundamentals**, enabling informed architectural decisions and credible engagement with data science teams.* Leverages **PyTorch and TensorFlow** for embeddings, training pipelines, and targeted fine‐tuning where business value is clear.* Ensures GenAI capabilities integrate seamlessly into the broader **ML, data, and MLOps ecosystem**.* Balances rapid GenAI delivery with long‐term model sustainability and governance.**6. Production Deployment, Scalability & Operational Excellence*** Leads deployment of GenAI systems into **secure, scalable production environments** using **Docker, cloud‐native architectures, and hardened APIs**.* Establishes **observability and monitoring** for GenAI applications, covering performance, drift, quality, reliability, and failure modes.* Ensures GenAI platforms meet **enterprise availability, resilience, and disaster recovery expectations**.* Drives operational readiness, incident management, and ongoing optimization of AI services.**7. Software Engineering Leadership*** Brings strong **hands‐on software engineering credibility**, setting standards for Python‐based GenAI services.* Leads development of **high‐performance AI‐powered APIs** using FastAPI and async programming patterns.* Champions clean architecture, testability, and security best practices across AI engineering teams.* Acts as a bridge between **traditional application engineering and AI‐native development**.**8. AI Safety, Evaluation & Responsible AI Governance*** Leads the implementation of **AI evaluation and governance frameworks**, including hallucination detection, confidence scoring, and human‐in‐the‐loop validation.* Designs and enforces **guardrails, moderation layers, and usage controls** to prevent misuse or unintended outcomes.* Partners with Risk, Compliance, Legal, and Security teams to embed **Responsible AI principles** into all GenAI solutions.* Ensures GenAI adoption withstands **audit, regulatory, and reputational scrutiny**.**9. Leadership, Influence & Execution*** Operates as a **hands‐on SVP**, combining strategic influence with deep technical execution.* Leads senior engineers and GenAI specialists, building **sustainable internal AI capability** rather than point solutions.* Communicates complex GenAI concepts clearly to **executive and non‐technical stakeholders**.* Drives delivery in **agile, fast‐moving environments**, with a strong bias for outcomes and measurable value.**Recommended Qualifications:*** **10+ years of progressive experience** in software engineering, ML, or AI platforms, with **5+ years leading senior engineers and architects**.* **3+ years of hands‐on experience deploying LLM‐based systems** in production environments at enterprise scale.* Demonstrated authority across **commercial and open‐source LLM ecosystems** (e.g., OpenAI, Anthropic, Google, Llama), including model selection, fine‐tuning, and hosting strategies.* Proven ability to define **enterprise-wide GenAI standards**, reference architectures, and reusable accelerators.* Demonstrated leadership in establishing **prompt engineering standards and orchestration patterns**.* Experience optimizing **latency, throughput, accuracy, and token cost** across large‐scale GenAI workloads.**Education:*** Bachelor’s degree/University degree or equivalent experience* Master’s degree preferred