Is your Data Architecture Optimized for your Front, Middle and Back Office?
Explore the following topics:
Learn what leading brokerages and financial institutions are doing to improve their customer experience and risk management by enabling real-time performance across large customer bases and data sets.
With the surge of trading volume by both new and experienced traders, it is both good news for online brokerages with the volume of business but it comes with distinct challenges to the underpinning data architecture.
Your data infrastructure has to support an increased reliance on mobile, provide competitive advantage for profitable margin lending, and facilitate greater customer personalization for upsell, cross-sell and retention, all while keeping fraud in check.
How Front, Middle, and Back Office apps can benefit from a Better Data Architecture
Brokerages have vast amounts of data in legacy systems and systems of record that need to be integrated with modern applications such as credit risk, fraud management, client-facing services, or transaction posting.
To help customers with recommendations to optimize their portfolios, for robo advising, or for sales and research staff, brokerages should look to leverage massive, accurate data sets and conduct large scale analysis with cutting edge artificial intelligence and machine learning (AI/ML) – all in real-time.
To drive profitable margin loads, drive decision support (authorizing trades), conduct compliance (both pre- and post-trade), and conduct performance measurement (attribution algorithms/modeling), firms will need to cost-effectively make customer portfolio data available from their system of record available in a consumable format, and also in real-time.
For back office, a better, real-time data architecture will help:
- Supply accurate, complete data to market makers
- Track commissions and charges
- Maintain custody of securities
- Automatically execute settlement and reconciliation
- Conduct funds administration
- Minimize fraud exposure
- Deliver risk reporting in real-time
To learn more about this and more, watch the discussion.
Leveraging a Better Data Architecture for AI/ML-powered Brokerage apps
TAI/ML applications have insatiable appetites for data, and the portfolio optimization, margin lending, risk assessment, post-trade compliance are no exception. Machine learning models run better with more data, and the more iterations and the more training, tuning and validation you can do, the better your results. The challenges lie in data preparation (which is painful) and model creation and tuning as models are constantly evolving. Plus, you need an online system with streaming data and the need to make an inference in milliseconds. The problem is wanting to pull disparate signal data from sources from different countries and data centers in real-time.
Aerospike feeds hungry ML systems more data, faster and efficiently with our real-time data platform that:>
- Cuts Spark-based jobs execution time by 80%>
- Reduces ML model training times>
- Increase frequency of ML model retraining>
- Enables in-situ data exploration eliminating compliance headaches by removing the need to copy data into multiple systems>
- Creates a low latency inference or online training pipeline>
Predicting the Future in Trading; 80TB at a Time with Aerospike
A leading quantitative research company was looking to leverage massive data sets and large scale analysis with cutting edge machine learning for quantitative trading advantages.
Prior to adopting Aerospike, the firm was using SQL Server + caching layer but they were unable to get as much data as they’d like, nor was it as current as they felt it could be. With Aerospike, they were able to keep up to 20 years of data online and update it every five minutes and have predictable expenses around data storage growth needs.
The firm was looking to be able to optimize their data architecture, enabling them to refine their financial models. With Aerospike, the firm is able to:
- Actively monitor 200k products
- Track 200 fields for each product
- Maintain up to 10 versions of each data point
- Conduct updates on every data point every 5 minutes
- Track 20 years of data online
“Data refreshes that hadn’t been possible due to length of time needed, are now possible to complete in 12 minutes on Aerospike ..leading to data that is 10-20% more accurate.”
Vice President, Engineering
Leading Quantitative Research Firm
How one of the World’s Largest Brokerages Benefits from Aerospike
One of the Global Top 3 Brokerage’s IT group was at a crossroads. The combination of an RDBMS and a RAM-based cache fronting a traditional mainframe database was unable to consistently and reliably support ever-growing workloads during trading hours. This reality had become a major impediment to the company’s strategic goal of releasing a steady flow of new and updated applications to its mobile customer base.
In deploying Aerospike, the company:
- Performs 250 million transactions and 2 million price updates per day
- Saves $10,000 per day with mainframe offloading – a 90 percent reduction in TCO
- Reduced server count from 150 down to just 12
- Makes real-time customer portfolio data available to drive margin loans
- Implemented a new intraday operational data store
- Accelerated new, competitive capabilities by providing easy access to vast amounts of data
“We’ve not had issues where customers complained about something not processed correctly on the backend system. It’s a true reflection of the capabilities that Aerospike provides.”
Vice President, Technology
Top 3 Global Brokerage Firm