Lead Data Engineer, Analytic Platform

Get to Know the Team 

The Integrity team is dedicated to Trust, Identity, and Safety, safeguarding our users & transactions on Grab. We leverage our expansive datasets to address critical challenges, protecting millions of users from sophisticated fraud, systemic identity abuse, and severe safety violations. Today, we are undertaking an aggressive transformation to become a "CybOrg"—moving away from manual data analysis towards an autonomous state where highly resilient, autonomous agentic workflows redefine how we detect and mitigate risk at scale. We rely on our technical builders to architect the advanced data infrastructure and orchestration layers that give our AI agents their "brains" and keep Grab ahead of evolving threats.

Get to Know the Role 

Reporting to the Head of Analytics for Risk, you will serve as the critical "Agentic Middleware" layer for the Integrity team. This is not a traditional analytics or engineer role; you will be the technical architect connecting our Trust & Safety business logic to our real-time Risk Engine. You will help the team to build the data pipelines, vector databases, and multi-agent orchestration frameworks that power our AI. You will solve deep backend technical bottlenecks, build proactive guardrails, and ensure our autonomous systems can operate reliably and accurately without breaking core production systems.

The Critical Tasks You Will Perform

• Architect the AI Data Infrastructure: Build and optimize the data ingestion pipelines behind our AI, including vector databases, embeddings generation, and parsing unstructured data (such as Slack logs and historical investigations) to enrich the agent's knowledge base.
• Orchestrate Multi-Agent Workflows: Design, deploy, and maintain end-to-end multi-agent systems using modern orchestration frameworks (e.g., LangChain, LangSmith, LangGraph) to automate complex risk detection workflows.
• Build Proactive Blast-Radius Controls: Implement automated, code-level intercepts (such as Abstract Syntax Tree (AST) parsers and EXPLAIN query cost evaluations) to block hallucinated or destructive AI-generated SQL/rules before they hit production.
• Develop Secure Deployment Guardrails: Build tenant-owned Model Context Protocol (MCP) servers and highly restricted API wrappers that allow AI agents to safely interact with Grab’s core Risk Engine without risking platform stability.
• Optimize System Reliability: Own the technical stability, latency optimization (e.g., resolving deep asyncio blocking bugs), and token-cost efficiency of our internal agent platforms so our Data Analysts can focus purely on the Risk domain
• Collaborate as a Systems Fusion Expert: Partner closely with domain-expert Data Analysts to translate their tacit Trust & Safety knowledge into structured, machine-readable formats and scalable agentic skills.