DispatchGym: Grab’s reinforcement learning research framework

Summary

Applying reinforcement learning to dispatch systems is often hindered when the chosen control levers exert weak influence over reward functions. To streamline research, Grab built DispatchGym, a framework that connects reinforcement learning algorithms to a dispatch process simulation via the Gymnasium API. The simulation emphasizes directional accuracy over absolute precision, allowing researchers to evaluate relative metric shifts across supply and demand scenarios. Built in modular Python and accelerated with Numba, the system allows data scientists to test code locally and launch distributed Spark executions with a single command-line call. The framework has been used to evaluate various contextual bandit models and action sampling strategies for tuning dispatch hyperparameters.

Takeaways

• Reinforcement learning becomes ineffective when a manipulated lever exerts negligible influence on the reward function, making reward sensitivity and lever selection critical.
• DispatchGym's simulated environment prioritizes directional accuracy over absolute real-world metric matching, ensuring reliable sim-to-sim comparisons during research.
• To lower barriers while preserving efficiency, DispatchGym combines a modular Python codebase with Numba acceleration and runs distributed experiments as Spark jobs via a single CLI command.

Related reading

• Grab · Sep 23, 2025

  Powering Partner Gateway metrics with Apache Pinot

  Grab needed to power real-time analytics dashboards for its Partner Gateway, tracking API status codes and latency across datasets reaching 6.8 billion rows over 30-day windows. Initial aggregation queries on datasets exceeding 150GB frequently timed out past 10 seconds, failing to meet the platform's 300-millisecond service level agreement. To support low-latency Online Analytical Processing queries, Grab routed metric streams through Apache Kafka and Apache Flink into Apache Pinot. Query execution was then accelerated by partitioning Kafka topics by metric name, adding rounded time interval columns, and implementing Star-tree indexes for multidimensional pre-aggregation.

  Alvis Chew

• Grab · Jan 29, 2026

  Cursor at Grab: Adoption and impact

  Following a multi-tool AI strategy, Grab integrated the AI coding assistant Cursor into its engineering toolkit in late 2024 to accelerate software development. Technical staff adoption reached 98% monthly active usage with a 50% suggestion acceptance rate, supported by custom monorepo indexing and preconfigured rules aligned with internal coding conventions. Engineers frequently apply the tool to unit test generation, code refactoring, cross-repository navigation, and routine API scaffolding, with over a third of merge requests incorporating Cursor. The rollout also encompasses non-technical personnel and product designers who, after receiving Git training, submit direct production UI fixes. Statistical evaluations using fixed-effects regression indicate a dose-response relationship between Cursor usage intensity and measurable productivity gains.

  Akshay Misra

• Grab · Jan 21, 2026

  Docker lazy loading at Grab: Accelerating container startup times

  Grab addressed slow cold starts and auto-scaling bottlenecks caused by large container images across data platforms like Airflow and Spark Connect. To eliminate the requirement of downloading complete images before launching, the team evaluated Docker lazy loading via remote snapshotters using eStargz and Seekable OCI (SOCI). Unlike eStargz, which modifies image layers and increased application startup delays during benchmarks, SOCI stores index metadata separately as OCI Artifacts without altering image digests. In production on Amazon EKS, tuning SOCI concurrency and chunk parameters reduced fresh-node image download times by 60%, ultimately delivering a 30% to 40% reduction in P95 startup times.

  Huong Vuong

• Grab · Nov 21, 2025

  SpellVault’s evolution: Beyond LLM apps, towards the agentic future

  Grab developed SpellVault as an internal no-code platform to democratize the creation of AI applications backed by Retrieval-Augmented Generation (RAG) and plugin integrations. To advance beyond static retrieval and linear input-output processing, the platform transitioned from its legacy executor to a graph-based execution model supporting branching, looping, and ReAct agent patterns. Capabilities like Python code execution and internal repository searching were unbundled from the prompt builder and consolidated alongside user plugins into unified Native and Community Built Tools. The platform also introduced a drag-and-drop deterministic workflow designer, automated task scheduling, and support for the Model Context Protocol (MCP).

  Felix Haryanto Lie