Transwarp Astro
Intelligent Data Governance Platform
Product Introduction
Transwarp’s Astro Intelligent Data Governance Platform uses an agent orchestration approach to fully automate data development, governance, querying, and analytics—all from a single natural-language request. At its core, Astro makes conversational collaboration the centerpiece of the product experience, delivering three key capabilities: proactive clarification—it asks follow-up questions when business requirements are unclear rather than guessing an answer; transparent planning—it keeps multi-step task plans fully visible and adjustable on the fly; and instant replanning—it lets the system correct itself immediately when a user simply points out. Built on these three capabilities, Astro already offers 16 task templates across three common enterprise scenarios, creating a complete closed-loop workflow that supports 24 different data roles and brings conversational agent collaboration into real-world production environments.
Core Features
Data Development
By orchestrating four intelligent agents, Astro automatically turns a single-sentence business request into a complete end-to-end data development workflow—covering everything from data asset discovery and data lake ingestion to multi-layer code generation and workflow orchestration. It starts by semantically matching the optimal data sources, then handles data lake ingestion and ETL task configuration in just minutes. From there, it automatically plans the entire data pipeline—from ODS (Operational Data Store) all the way to ADS (Application Data Service) — and generates both DDL (Data Definition Language) and ETL code. Finally, through conversational workflow orchestration and lineage-based dependency analysis, development tasks that used to take hours or even days can now be completed in minutes.
ata Query & Analytics
With two dedicated agents—Excel Analytics and Semantic Completion—plus reusable data asset discovery capabilities, Astro removes the major bottlenecks in data query and analytics workflows: understanding the data and doing it efficiently. The Semantic Completion Agent automatically generates human-readable annotations with built-in reasoning logic across six semantic dimensions, including field names, table names, and data samples. This lets undocumented tables effectively speak human language. Meanwhile, the Excel Analytics Agent lets users upload Excel or CSV files, automatically maps column names through natural-language interaction, and generates SQL for ad‑hoc analysis—returning results in minutes. Together, these capabilities allow analysts to go from raw data to visual insights in seconds, without relying on technical documentation or manual annotations.
Data Governance
Through the collaboration of seven governance agents, Astro delivers fully automated, end-to-end data governance workflows—from classification all the way to data asset publishing. Data Security Agent automatically determines asset classification levels using a four-step decision process based on lineage, knowledge bases, and related metadata. Data Standards Agent leverages LLM semantic understanding to standardize fields at scale and push enforcement rules directly into execution. Data Quality Agent generates end-to-end quality rules covering six dimensions of data quality. Data Lineage Agent supports field‑level lineage parsing across multiple SQL sources and ETL tools, while building a global lineage topology. Impact Analysis Agent automatically traces lineage dependencies up to five levels deep to assess change risks. Asset Tagging Agent intelligently recommends and applies business tags based on the enterprise's taxonomy. Asset Cataloging Agent automatically matches catalog structures, batch‑mounts data tables, and triggers asset publishing. The entire process runs with minimal manual intervention, dramatically improving governance efficiency.
Advantages
Conversational Data Governance Agent Collaboration
With a single natural-language request, users can orchestrate the entire lifecycle of data development, governance, querying, and analytics—turning multi-platform manual work into a conversational closed loop and making the platform far more accessible.
Dynamic Multi-Agent Orchestration
Based on user requirements, the orchestration engine automatically plans execution workflows, coordinates multiple agents in parallel or sequence as needed, and intelligently bypasses completed steps—all to maximize overall task efficiency.
Proactive Clarification and Transparent Planning
When faced with ambiguous business requirements, the system proactively asks clarifying questions to ensure alignment on understanding and definitions. Execution plans remain fully visible, editable, and extendable throughout the entire process, giving users complete control and full transparency over the workflow.
Feedback and Replanning Mechanism
When results fall short of expectations, users can simply correct the system using a natural-language instruction—automatically triggering replanning and self-correction. This boosts both system robustness and usability.
Application Scenarios
Data Development Scenarios
Astro transforms the traditionally slow and fragmented data development process. A request like 'monthly contract value trends by contract type' once required developers to navigate 4 systems, complete 7 separate stages, and spend up to 4.5 days. Now, through a single natural‑language request, multi‑agent execution (both parallel and sequential), and one‑time review‑and‑publish, Astro automates requirement understanding, workflow planning, and coordination across asset discovery, data ingestion, code generation, and orchestration—reducing delivery time from days to minutes.
Data Governance Scenarios
Astro solves the 'module marathon' challenge common in traditional governance tools. In the past, when developers faced urgent tasks like 'govern and publish five customer tables,' they had to constantly jump between 5 separate modules—metadata management, data standardization, data quality, asset tagging, and data asset cataloging—often enduring up to 20 context switches. Governing a single table could take as long as 80 minutes, and managing 20 tables in batch meant the workload scaled linearly. Now, Astro delivers a one-stop governance closed loop driven by single-sentence requests, Excel template inputs, collaborative governance agents, and human-in-the-loop confirmation at key checkpoints. The system automatically recognizes user intent and orchestrates agents to handle semantic completion, data standardization, data quality management, asset tagging, and asset mounting in sequence. With built-in state awareness and intelligent step‑skipping capabilities, Astro significantly reduces governance time per table while enabling highly efficient large‑scale batch governance.
Data Query & Analytics Scenarios
Astro eliminates the inefficiencies in data analytics caused by missing metadata. Previously, when a new analyst received a seemingly simple request like 'search the monthly execution progress for the tb_contract_execution table,' they often had to spend two hours reviewing code, consulting colleagues, and hunting through documentation just to understand what each field meant—because only 2 out of 18 fields had proper annotations. As a result, a five‑minute query would turn into a hours‑long ordeal. Now, Astro delivers a conversational data analytics closed‑loop built on single‑sentence requests, intelligent agent selection, automatic semantic completion, and rapid insight generation. This dramatically lowers the barrier to querying data and compresses hours of field interpretation into seconds.
Customer Success Stories
A leading futures institution integrated 4 core business systems and implemented over 70 futures industry classification standards, along with more than 120 custom standards. The platform then performed intelligent classification and grading on over 4,000 data assets with 90% accuracy—significantly outperforming traditional rule‑based engines.
Transwarp, Shaping the Future Data World
