By Peter Simpson, OneTick Product Owner
In our recent Tech Update webinar, Mick Hittesdorf and I walked through the major enhancements in OneTick 1.26. This release represents the culmination of a year’s worth of development focused on three specific goals: making analytics faster, making the platform easier to use via SQL and Python, and ensuring interoperability through open standards.
Whether you are deploying on-premise or managing a complex cloud infrastructure, here are the key differences in capabilities, performance, and access you can expect with version 1.26.
1. Capabilities: A SQL and Python-First Approach
We are seeing a clear shift in how our customers interact with data. While our proprietary OneTick Query (OTQ) language remains powerful, the focus of 1.26 is to enable users to perform complex financial analytics using standard SQL and Python.
- Financial-Aware SQL: We have expanded our SQL syntax to natively handle financial concepts. You can now execute operations like VWAP, TWAP, linear regression, and rolling windows directly in SQL. Furthermore, our SQL is "symbology aware," allowing you to query via Bloomberg, FIGI, or RIC codes while handling corporate actions and continuous contracts seamlessly.
- AI-Powered Assistance: To help you navigate these capabilities, we’ve embedded an AI Query Assistant directly into our documentation. You can type natural language questions—like "How do I calculate a TWAP?"—and the system will generate the correct SQL or Python code for you.
- Federated Querying with Trino: A highlight of the release is our new Trino connector. This allows you to federate data, enabling you to query OneTick data alongside data in Snowflake or BigQuery without complex ETL processes.
2. Performance: Vector Compute and Columnar Speed
Speed is non-negotiable. In 1.26, we have moved from row-based compute to vector-based compute, processing blocks of rows at a time to significantly improve query performance.
- Partitioned Parquet: We have spent the last two years optimizing Partitioned Parquet, and it is now ready for production use. In fact, for queries retrieving a subset of fields, Parquet is now faster than our traditional proprietary OneTick archives. This performance boost is achieved by porting page indexing from Java to C++ to optimize read speeds.
- Benchmarking: We don't just guess at performance. We now run a suite of over 200 benchmark queries—covering tick-by-tick, bars, and MBBO—against every release to ensure we are consistently getting faster.
3. Access: Zero Lock-In and Open Standards
One of the most significant changes in 1.26 is our commitment to interoperability. We want you to be able to access your data with or without OneTick binaries.
- Open Data Formats: With Partitioned Parquet, data is stored in a hive-partitioned folder structure (by table and date). This allows you to access OneTick-generated data directly using external libraries like pyarrow without touching the OneTick server.
- Arrow Flight & ADBC: We have added Arrow Flight SQL support. Enabling it is as simple as adding a single server port line to your config. This allows connection via ADBC drivers, returning data as Arrow tables for high-performance conversion to DataFrames.
- Thin Client Connectivity: We have retired the need for heavy local installs for many users. You can now simply pip install onetick-py and connect via REST using Arrow as the data transport, which is now faster than the traditional fat client for many data retrieval tasks.
Housekeeping Notes
As we modernize, we are also retiring support for older technologies to maintain security and performance. With 1.26, we are dropping support for 32-bit Windows builds, older Visual Studio versions, and Tomcat 9 (moving to Tomcat 10.1).
OneTick 1.26 is about removing the "plumbing" work so you can focus on analytics. If you would like a personalized walkthrough of how these features fit your environment, please reach out to us.
Visit onetick.com, email info@onetick.com, or request a private demo here.
Contact us today to set up a personalized walkthrough of these new capabilities.
Best wishes,
Peter Simpson, OneTick Product Owner
