As of March 2009, there are 64 implementations of Velocity in production, in North America, Europe, and Asia. Clients include 8 of the top 10 tier-1 financial institutions and comprise major global investment banks, the largest online brokerage, highly regarded hedge funds, the world’s largest Commodity Trading Advisor, and proprietary trading groups at leading global financial institutions. Most installations are strategic market data infrastructure solutions supporting diverse applications on the same underlying platform and systems. We deliver the fastest market data solutions for the capture and high-speed analysis of massive amounts of streaming and historical data – and based on market and client/prospect feedback, we are unique in this specific value proposition.

The Velocity database is a proprietary time series store, optimized for size and performance. It is not relational, as this would impact the performance negatively. The powerful combination of our breakthrough technology in the areas of compression and advanced caching alleviate any concern regarding the ability to handle increasing volumes of market data.

Vhayu Velocity is an integrated solution that can store and analyze both real-time and historical data from the same engine. It manages in-memory and on-disk data seamlessly from the user's/query's perspective. The system uses high-performance indexing, caching, disk and memory look-up technologies to provide the optimal query performance even over large amounts of data.

Vhayu can support any data feed. Vhayu Velocity ships with out-of-the-box support for most major data vendors or direct exchange feeds. New feed handlers are easily developed for clients as needed on a professional services basis or by the customer via a feed handler SDK.

Velocity provides a flexible instrument listing and symbology mapping interface to control the data subscription from input sources. In addition, Velocity also handles broadcast feeds that allow access to all of the available instruments. The symbology functionality allows users to easily manage symbol changes – new symbols that are added and old symbols that are deleted. Velocity supports a symbology-mapping mechanism that allows an end-user to define any alternative symbology to be used for data queries. For the addition of new symbols, a change is required to get the upstream feed to provide the data for the new symbols to Velocity’s feed handler. This may require adding the symbol to a subscription list being maintained by the feed handler; this would be the case if the upstream feed is a per-stock subscription-based feed, as opposed to a wildcard/broadcast/fire-hose-based feed. Once the feed handler receives the data for the new symbol, the downstream components create an internal representation to track it automatically.

Velocity can support any asset class. We have clients in production using Velocity for equities, futures, FX and fixed income.

The Velocity Broadcast Servers can handle multiple concurrent data feeds from different feed vendors.

Velocity supports sub-microsecond time. With this flexibility, Velocity can take any timestamp from the exchange and simply store it with the data. The database key includes the symbol and time along with other necessary data for the record being stored. Velocity also stores additional “sequence numbers” allowing us to recreate a tick stream at a higher resolution than the time field given to it by the exchange/feed.

Vhayu Velocity runs dual/quad core or higher, 64-bit Windows 2003 servers offering the best price/performance ratio.

Velocity scales linearly thus enabling customers to scale efficiently, as the volume of data, complexity of analytics, and number of users increase. There is no limit to the number of users that can be supported by Velocity. The system scales quickly and economically by adding more processors and more RAM. More importantly, the Velocity architecture means more Analytic Engines can be added to expand system capacity without putting extra load on a customer’s market data infrastructure.

Vhayu Velocity is an enterprise solution with built-in redundancy and automatic failover throughout. We employ a primary and secondary set-up, making it possible for Vhayu Velocity to continue operation in the event of possible component failures. As an added measure of protection, disaster recovery sites can be set-up with a similar primary and secondary set-up. The disaster recovery site will collect and provide all tick and quote data as a back-up source. In addition, Vhayu Velocity can constantly monitor its input data sources and send out alerts when a feed source failure is detected. Vhayu Velocity has the ability to failover from one feed server to another.

Velocity offers both an on-demand query model and the event-driven model, to give maximum performance and flexibility to users. It uses extremely efficient multithreading, allowing the system to support very large numbers of users running simultaneous queries. Velocity's unique multithreaded database allows very high query performance even while capturing large volumes of streaming real-time data. We recognize that the querying performance while capturing real-time data during market hours is critical for our customers, and Vhayu is the recognized market leader in providing this capability through the use of multiple cutting-edge technologies.

Vhayu Velocity APIs allow for easy integration to any input source and to any client application. We recognize the need for flexibility and performance, and these concepts are inherent in the data access interfaces exposed by the system. Vhayu Velocity APIs enable clients to work with languages with which they are familiar, eliminating the need to learn proprietary languages.

Vhayu Velocity core API is exposed in two mediums:

On the client side, we have a scripting language that is easy to learn and uses a string-based, platform-independent, language-independent query mechanism that allows any application anywhere on the network to request data and/or analytics at very high speeds.

We also have client APIs exposed through high-level languages (such as C++ and Java), allowing users to write and extend applications using the language with which they are familiar.

SQL may also be used via ODBC/JDBC, which lets non-technical users query the database directly and enables integration with industry standard reporting tools.

On the server side, we have a programming interface to create libraries that can directly access the data and internal analytic API implementations.

In addition, we have native implementation for 3rd party applications such as R, MATLAB, S-PLUS and Excel. This allows users to continue using their favorite tools and bring in the data and results of analytics from the server for further analysis.

Velocity supports a client-side scripting language that follows the syntax of a language called Tcl, which is easy to learn and uses a string-based, platform-independent, language-independent query mechanism that allows any application anywhere on the network to request data and/or analytics at very high speeds.

In addition, Velocity supports SQL access via ODBC from Windows and Linux or JDBC for cross platform Java connectivity.

We also have client APIs exposed through high-level languages (such as C++ and Java), allowing users to write and extend applications using the language with which they are familiar.

Initial training is included with your Velocity license. Premium training services are available. Our approach is to have customer resources working closely with the Vhayu team at every stage of an implementation. User training is provided to customer team members at multiple stages of the implementation. This focused approach to transfer of knowledge and skills to the customer team makes the transition from implementation to production support easier. Vhayu can also provide professional consulting services to help clients integrate content, build analytics or implement the Velocity solution at one or more locations.

While there is no out-of-the-box FIX connectivity, we have integrated with order management systems in the past for customers. FIX connections have been made by integrating the QuickFix libraries and other OMS via their own proprietary API (e.g., Portware, FlexTrade, InfoReach).

Velocity supports a wide variety of industry standard communication and messaging protocols.

For communication, we support TCP/IP, UDP, Multicast, Tibco and more. For publish/subscribe, we support subscription over TCP/IP and publishing over TCP/IP, multicast, RV, RMDS and RFA.

Vhayu Velocity comes with interfaces to Excel (via our XLStream), R, MATLAB and S-PLUS.

Because most of our users are major brokers, hedge funds and alternative trading systems looking for the fastest performance available, coding in Velocity is done in C++ for optimal speed and flexibility. There are visualization products that come bundled with Velocity, including Vhayu Vision™ and an Excel interface called XLStream™. Our flexibility allows easy integration to existing client and other applications. All Vhayu interfaces are provided in all industry standard languages.

POCs with defined acceptance criteria on customers’ premises are available upon consultation with Vhayu. Vhayu can also provide demonstrations with notice.

Installation of Velocity is relatively quick, with time and effort depending on the chosen architecture. Consulting and implementation of client requirements can only be estimated once they have been exposed to Velocity and thoroughly analyzed. Recent case studies have shown that implementations closely aligned with POCs require 2 to 4 weeks.

Corrections and cancellations are applied to the data as soon as the record is received from the feed. The original record is retained, and by default the engine will apply cancels/corrects to the tick stream on the fly to present the ‘cooked’ stream. Users may query the clean dataset, which is the data prior to cancels or corrects.

The Velocity historical archives can be populated with proprietary data, TAQ data, Nexa TickData or Thomson Reuters DataScope Tick Histories. In addition, a visual GUI tool is provided to map other data sources into Velocity. Velocity is completely cross-asset and can store all item-based data from any source.

Use of fault-tolerant servers, a pair of fault-tolerant Broadcast Engines, and one or several feed loggers make the probability of data loss highly unlikely. Gaps in archives can be filled later on using historical tick data files.

Velocity has extensive support for corporate actions, including splits, reverse splits, stock dividends, cash dividends, mergers, de-mergers, etc. Velocity provides maximum flexibility and ensures historical data sanctity by retaining the original data as-is, and applying the corporate actions at very high speeds on-the-fly.

Velocity has built-in capabilities to handle exchange holidays and weekends, and runs 24x7 without system interruption.

Benchmarking a real-time analytic engine requires a precise definition of the endpoints of the system and components being benchmarked. Practically, a market data analytic system will include telecommunication elements, network hardware/software elements, feed software/hardware elements, feed handler libraries, feed analysis components, result distribution components, and business applications/systems.

Please note that Vhayu will be happy to arrange for specific latency benchmarks to be implemented in a well-defined environment. Velocity’s high-performance thread synchronization technologies, data schema, and event management technology differentiate us from solutions lacking the same. Since application-managed event and thread synchronization do not have to resort to kernel/OS layers as heavily, the CPU cycles saved during high-performance parallel reads/writes enable Vhayu Velocity to deliver superior real-time parallel query/event processing performance.