Senior AI Engineer

Role Overview

• Develop an understanding of Grabs Financial Services and Banking data and its nuances across countries.
• Deploy enterprise-grade agentic solutions that power the next generation of customer experiences.
• Develop AI solutions for use cases such as representation learning, conversational AI and agentic operations.
• Fine-tune and adapt LLMs using techniques such as supervised fine-tuning, instruction tuning, LoRa and PEFT.
• Build transformer-based models from scratch, including foundational models and promoting models trained on large-scale transactional data.
• Manage the complete lifecycle of AI systems from experimentation and evaluation through deployment, optimization, monitoring, and maintenance.
• Work with engineering, product, compliance, risk, and operations teams to build reliable and responsible AI solutions.
• Stay current with AI and LLM research, frameworks, and new techniques, and translate them into practical applications.
• Or collaboratively to solve complex and ambiguous technical problems.
• Operate as an individual contributor with 5+ years of relevant experience.

Daily Activities

• Enterprise Agentic AI: Design and deploy production-grade agentic systems using LLM orchestration, tool use, planning, memory, retrieval, and multi-agent workflows. Build for reliability, observability, security, scalability, and human oversight.
• LLM Development and Fine-Tuning: Fine-tune and evaluate LLMs using techniques such as supervised fine-tuning, instruction tuning, LoRA, and PEFT. Build high-quality training and evaluation datasets and increase models for accuracy, latency and cost.
• Transformer and Foundation Model Development: Design and implement transformer architectures from scratch. Train foundational or domain-specific representation and promoting models using large-scale transactional and behavioural data.
• Evaluation and Guardrails: Define offline and online evaluation frameworks for AI agents and models. Develop evaluation datasets, safety guardrails, monitoring systems and feedback loops to improve production performance.
• Ownership and Collaboration: Design AI solutions by working backwards from business and customer needs. Lead delivery from problem definition and architecture through production deployment and ongoing operation.

Adherence to clean coding, modular system design, testing, APIs, and version control (Git/GitHub). Balance quality with production considerations such as latency, cost, reliability, security, and scalability.

Qualifications

The Must Haves

• Experience designing and deploying at least one enterprise-grade agentic solution in production, with practical experience in tool integration, orchestration, state and memory management, evaluation, guardrails, observability and failure handling.
• I have hands-on experience with agentic frameworks such as LangGraph, LangSmith, and LangChain. Additionally, I have a understanding of retrieval-augmented generation, LLM orchestration, tool calling, and agent evaluation.
• Transformer Architecture: understanding of transformer architectures, including attention mechanisms, tokenization, positional encoding, training goals, optimization, and inference. You can implement and train a transformer-based model from scratch
• Core ML: Expert proficiency in Python, Spark, and SQL (Presto/Hive) with a command of fundamental ML concepts (Bagging, Boosting, Recommendation Engines)
• AI Productionization: Expertise in deploying scalable AI and ML systems using tools such as MLflow, Kubeflow, TFX, or SageMaker, together with experience in model serving, monitoring, automated evaluation and cloud infrastructure.

Disclaimer: This job posting has been aggregated from external source. Role details, content, and availability are subject to change. Applicants are advised to confirm the latest information directly on the company website before applying.