Advanced Heuristics &
Intelligent, Learning, Adaptive
A I H E U R I S T I C S T H A T L E A R N & A D A P T
Technology & Engineering ...
Quantitative Technologies pioneered Autonomous Black-Box Trading Systems based on the latest Third Generation Adaptive Algorithmic Trading methodology as well as Artificial Intelligences in medium and high-frequency trading technology that are capable of self-healing, self-monitoring and self-defense. These employ various switching methodologies among multiple trading strategies in a portfolio of systems mostly in currencies, futures and equities based on multiple integrated artificial intelligences.
Technologies and algorithmic systems licensed and used by banks, major hedge funds (under non-disclosure).
Multiple award winner in Artificial Intelligence from Ford, AT&T, Ericsson, Samsung, Amazon, Intel and also worked with Deutsche Bank, Citibank among others.
Numerous issued patents in algorithmic application which are licensed and used by every major bank.
Pioneered artificial intelligence in analyzing marketing data using similar approaches for analyzing markets for predictive marketing analytics to learn what content drives customer interaction.
Industry-Leading Technology Innovation Awards in Artificial Intelligence...
Over the years we have been issued numerous patents for the technologies we have developed. Many have been sold to Microsoft and licensed to most major banks. But the most rewarding has been the numerous awards we have been awarded from some of the largest Fortune 100 companies including Ford, Intel, Amazon, Samsung, Ericsson, among others.
Data & Computing Infrastructure
We maintain arrays of servers and data clusters colocated on the same trunks as the Department of Defense for resiliency and speed. In addition, we are located next to one of the few Super NAPs on the planet. The Super NAP we chose is Tier IV Gold, and is also used by the Department of Defense and other military agencies.
We maintain continuous analysis of most major markets for many of our clients, and offer those same data analysis and artificial intelligence deep analysis for lease. If you require detailed technical or fundamental market analysis, we can schedule analytics cycles for your tasks as well.
CONTACT US: To discuss your needs and pricing.
The heart of any intelligence is the management system. Our trading systems use advanced algorithmic based, portfolio optimization technology that borrows from the latest academic research as well as traditional algorithmic and machine learning approaches. Our difference is that all the algorithms must compete against each other.
So each individual algorithm is constantly competing to show that they can generate more alpha from their management approach, better than their sibling algorithms. Then overseeing all the algorithms is the Omniscience Engine Artificial Intelligence. This AI sees all using advanced heuristics, and make the final determination based on which algorithms are generating the most alpha and profits. Then the Omniscience Engine AI constantly re-balances among the different algorithms using third-generation algorithmic portfolio rebalancing methodology to constantly shift the return curve towards optimal. The net result, is a smooth, consistent, and adaptive alpha curve that optimizes for consistency, and risk control above all else. These systems are custom designed per client so as to not collide with other trading systems.
Cortical Arrays & Algorithmic Development
The challenge that most Hedge Funds and Private Equity firms face is the development of new algorithms, especially in rapidly shifting market regime changes. Existing algorithms quickly lose their "edge", and developing new algorithms is time consuming, and cannot react quick enough with requirements of forward testing.
The Quantitative Cortical Array is an array of hundreds of artificial intelligences working in unison, with each AI focusing on a particular market, and a particular aspect of that market. Each AI utilizes machine-learning-heuristics to search for inefficiencies that can be harvested for profit. All these AI's do so in real-time.
In addition, each AI spawns children. Each child splinters the aspect view of the market, further fine tuning, and at the same time searching for non-correlating inefficiencies for harvest. This leads to literally thousands of algorithms exploring hundreds of markets, and thousands of equities for inefficiencies.
The net result is real-time, and very rapid discovery of opportunistic AI based algorithms capable of producing solid returns, and quickly adapting to whatever market conditions change, in a matter of seconds.
This approach works very well on concert with the Omniscience Artificial Intelligence Engine to server as a management algorithm to all the other AI's and their offspring.
Swarm Optimized Liquidity Solutions
The problem with institutional trading is the constant trade-off between scale, and liquidity.
To solve this problem, here at Quantitative Technologies, we approached the problem from a different perspective, using the biomimetic approach of swarm optimization. In essence, splintering a single trade in to hundreds, and thousands of execution components. The result is a collection of "trades" or executions that are now in essence a swarm of executions.
This was originally done to solve the problem of limited liquidity. By splitting a single trade into a swarm of smaller trades, you could then execute at much higher liquidity with minimal impact on net fill pricing and hence performance.
Now you have a collection of executions that are each given their own intelligence (AI). So each AI becomes a member of the swarm, with the intention of executing in the market at the most optimal price. In the same way that millions of birds cannot fly through a tunnel at the same time, thousands of executions cannot happen at the same time without colliding with liquidity issues.
With swarm optimization, each execution thinks and behaves independently. Imagine tens of thousands of bats, flying out of a single hole in a cave. Each bat arranges it's location, and exit based on the bats around it, and the exit. In addition, each bat is communicating it position, location, trajectory to all the other bats. These AI's behave the same way, communicating with each other, and finding the most optimal time and size to enter, or exit the market.
The net result is the ability to execute in limited liquidity with higher orders of magnitude.
Advanced Swarm Optimized Alpha Systems
One of the most powerful techniques that we have discovered, and has provided a strong, consistent alpha in all market conditions, has been an advanced form of swarm optimization. By combining swarm optimization and hive intelligence, has allowed hundreds, and thousands of individual algorithms to use seek and hunt patterns to search out different markets, different approaches among different time horizons, each in search of alpha, and inefficiencies that can be harvested.
The use of Hive Intelligence has made the swarm incredibly effective and efficient. By enabling all the individual drone algorithms to communicate with each other, they are able to coordinate themselves. So when one set of algorithms finds profits, it notifies other algorithms to join them in their search patterns.
The net result is a system, or hive, that can quickly adapt to market conditions, and more so, can actually profit from volatility through close coordination, and actually generate more alpha than would normally be harvestable.
Hive Intelligence Biomimetic System Design
Everything we do involved, massively scaled artificial intelligences. To solve the problem of inordinate computing power requirements, doubled with the highly dynamic requirements, of sometimes being mostly idle, to sudden market moves requiring exponential computing requirements, we combined biomimetic hive intelligence with a queen intelligence. It proved to be shockingly adaptable, and able to efficiently allocate computing resources beyond our wildest hopes.
In essence, each computing unit, or server is considered a drone. As a drone, it is capable of doing any task requested of it by the queen. Therefore, each server is adaptable, at a moments notice. The queen computing unit is capable of spawning (spinning up additional servers) at will.
So the queen, being aware of all of her drones, can allocate computing resources as needed. Under times of duress (or learned times of expected duress), she can request that additional drones be spawned in anticipation, and then released when no longer needed.
The system is highly adaptable, and incredibly resilient. Failures of single servers have no meaning, since other drones are always available to accomplish tasks. Finally, drones are capable of mutating into "daughter" servers, which can become queens if the queen becomes overtaxed, or damaged.
This highly adaptable, and nearly impervious approach is incorporated into all our engineering and infrastructure designs.
We are specialists, and pioneered Autonomous Black-Box Trading Systems based on the latest Third Generation Adaptive Algorithmic Trading methodology utilizing Artificial Intelligences in medium and high-frequency trading technology.
2831 St Rose Parkway
Las Vegas, Nevada 89138