New capability unifies in-flight data preparation with AI agent development, cutting AI development time
SAN FRANCISCO, Feb. 24, 2026 (GLOBE NEWSWIRE) -- The leading cause of AI project failure is data preparation bottlenecks created by disconnected tools. Today, Tray.ai, the leader in enterprise orchestration for data and AI, announced Tray Data Engineering, a new solution that combines data transformation, AI, and agent development to eliminate the delays that cause 60% of enterprise AI projects to fail.
As enterprises race to deploy agents, shift from pilot projects to production, and scale, they face the reality that data integration and AI stacks remain deeply disconnected. The result is AI supply chain gridlock, with AI and agent projects starved of data, throttling success. Pre-AI legacy data integration tooling is too unwieldy and inflexible to support the agile pipelines AI requires, and it is often siloed from agent development, causing business and AI teams to stall while waiting for data. Teams are forced to patch together separate agent development and integration tools, scripts, and legacy tooling, which chokes data flows, increases overhead, and creates technical debt.
Without access to timely, contextual, structured, and trusted data, agents and other AI projects fail. According to Gartner, "Through 2026, organizations that don't enable and support their AI use cases through an AI-ready data practice will see over 60% of AI projects fail to deliver on business SLAs and be abandoned."
At the foundation of Data Engineering is the new Tray SQL Transformer. Teams can execute sophisticated data preparation tasks directly within their Tray workflows, reshaping, joining, and transforming data in flight using a built-in, high-performance database and direct SQL-based transformations.
Because Tray is a unified platform that includes both integration and agent development, teams can manage the entire data-to-intelligence pipeline in one system, enabling enterprises to unlock significant efficiencies by consolidating legacy integration investments. Engineered data streams can power Tray Agents and MCP Tools, or be pushed to major data platforms such as Snowflake, Databricks, BigQuery, or Redshift via native Tray connectors.
"Data is the arbiter of AI success, but enterprises are gridlocked wrestling with disconnected, legacy integration tools that crush productivity and jeopardize results," said Rich Waldron, Co-founder and CEO of Tray.ai. "Tray Data Engineering increases efficiency in turning data into intelligence and is the secret to achieving AI outcomes at scale."
Key capabilities include:
Native in-flight SQL transformation: A new SQL Transformer provides a built-in virtual database so teams can create data pipelines to ingest millions of records (from formats such as Parquet and JSON) and use ANSI SQL to transform data and join across multiple files, all in flight. This eliminates the need for scripts, pre-processing, or bolt-on databases to make data AI-ready.
In-flight data hygiene and deduplication: To ensure AI models use only high-quality data, Data Engineering enables advanced cleansing operations in flight using standard ANSI SQL. Teams can easily deduplicate records, standardize text casing, and validate formats before the data ever enters a data warehouse or database, significantly reducing "hallucinations" caused by poor data quality.
Standards-based data transformation: Old-school integration often requires teams to use proprietary step-by-step logic, or aged-out languages like Ruby. The Tray platform now includes JSONata Inline Functions. This enables a modern and flexible approach to handling complex data transformations across any data type directly within Tray's visual workflows.
Flexible dual-mode output: Designed for maximum flexibility in the "last mile" of data delivery, transformed data can be output as JSON objects for immediate use in operational APIs and AI agents, or written to files (CSV, Parquet, etc.) for efficient bulk loading into major data lakes and data warehouses such as Snowflake, Databricks, BigQuery and Redshift.
One-step processing for cost predictability: Unlike legacy tools that force row-by-row iteration-which drives up latency, requiring deployment of workers and containers, or burning valuable task credits-Tray allows teams to process, join, and aggregate millions of records in a single database operation, powered by Tray's serverless architecture. This ensures pipelines are built quickly, and costs remain predictable.
Tray Data Engineering addresses the leading reason enterprise AI projects stall: disconnected, legacy data integration tools that leave AI projects started of data they need to function. By combining native SQL transformation, in-flight data cleansing, and AI agent development in a single unified platform, Tray eliminates the fragmented toolchains that create AI supply chain gridlock. The result is a streamlined path from raw data to production-ready intelligence-cutting AI development time , reducing costs, and providing enterprises with the data foundation needed to move confidently from AI pilots to scale.
Learn more at Tray.ai
Dive deeper into why AI agents are exposing the limits of legacy data pipelines
Register for the live webinar: Unblocking the AI data supply chain bottleneck
Gartner: A Journey Guide to Deliver AI Success Through AI-Ready Data, 11 July 2025, Ehtisham Zaidi, Roxane Edjlali
GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved.
Gartner does not endorse any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner's research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.
About Tray.ai
Tray.ai is the leader in enterprise orchestration for data and AI that enterprises use to build smart, secure AI agents at scale. It eliminates the need for disparate tools and technologies to integrate and automate sophisticated internal and external business processes and speeds the creation and deployment of high-value, production-ready AI agents. Enterprises can now avoid the traps of high costs and long lead times typical in custom agent development as well as the constraints and silos created by implementing and managing single-purpose agent offers from each SaaS application in the enterprise tech stack. With Tray.ai, the development of integrations, the delivery of intelligent apps and the integration of trusted data anywhere is fast, flexible and safe. Learn more at Tray.ai.
Media Contact:
trayaiPR@watersagency.com
