Prashant Bhuyan, CEO and Founder of Accrete.AI was cordially invited to speak to Cheddar.TV on the 'Between Bells' segment brought to viewers by IBM, to discuss the way AI transforms the way we work. Kristen Scholer, begins the interview directly,
"What are the biggest problems Accrete.AI is trying to solve?"
Accrete.AI is aiming to solve specific problems for money managers, investors are drowning in information and searching through all this information for actionable insights that might move markets is exhausting, and takes a lot of mental energy and the human brain just isn't equipped. The consequence of this is that investors are making bias decisions that lead to underperformance. Accrete's mission is to help investors overcome biases, make better decisions, and generate alpha in the digital age of markets. In particular Accrete is a market place of cognitive investment tools, automating cognitive tasks for domain specific real world problems related to information overload in the financial markets. Our tools read and understand huge volumes of information, and extract actionable insights and push those insights to customers via customizable watchlists, filters and notification. Empowering portfolio managers, traders, and analysts to spend less time searching for information and more time making well informed decisions. Baker Machado, presenter of Cheddar.TV asked,
"How do you use machine learning to potential help investment bankers and managers make those informed decisions?"
One of the challenges when you are trying to automate cognitive tasks is that you have to remove human bias and bias from machines, and make machines adaptive to change in contexts. One of the hurdles is that you don't always have enough relevant training data to teach the system 'How to learn' in a particular domain on a particular context. That gap is bridged by leveraging human expert networks together with artificial neural networks to create a feedback mechanism that is continuously de-biasing information as we ingest it, and improving the contextual awareness of the system all for the purpose of improving the accuracy of our systems. The end result is that we are able to create insights in a scalable way that are more accurate than humans or machines could do independently.
"How is IBM helping Accrete?"
Accrete.AI is backed by a strategic partnership that provides computational infrastructure, secure cloud computing as well as an enterprise grade distribution network via IBM's financial cloud services API marketplace.
"How does Rumor Hound and Topic Deltas work?"
Rumor Hound is an incredibly exciting product, as it solves a problem faced by many analysts, traders, and portfolio managers who are trying to keep ahead of rumors in the mergers and acquisitions space. There are a lot of rumors, they move the markets around, and M&A rumors can emanate from countless numbers of digital data sources, which would be impossible for even an army of human analysts to continue to read every chat room, blog, social media post or news article looking for actionable M&A rumors that might move the market. Rumor Hound does this for our users, it continuously reading, understanding, and looking for M&A rumors and all the linguistic nuances. When it finds a rumor it pushes this insight out to the user along with a credibility score, based on Alexa rankings and web demographic data, as well as a source URL and the rumor snippet itself.This allow users to validate for themselves whether the rumor is interesting. In addition, we are finding really interesting patterns in the data because we are seeing when low credibility rumors start to get picked up by higher credible sources, markets move! For example, December 2017 Target ($TGT) was an acquisition rumor by Amazon ($AMZN) which was from an obscure source which was later picked up by more credible sources, Street Insider and CNBC, and the stock popped 12% in five days. It seems like our users are making a lot of money and we are actually finding predictive insights, Accrete.AI gives reproducible research to customers so they can validate our findings.
"Are there things that have impacted the type of information Accrete provides?"
Accrete's core mandate is to process and contextualize natural language, unstructured data in real time so that our clients can overcome bias, make better decisions and generate alpha. The asset class doesn't really matter to us, whether it is crypto, currencies, equities, etc. What Accrete.Ai sees is a way for our customers to use our technologies to help people interested in a particular asset class, like crypto parse out some of the hype and the noise so they are able to get to the root of the problem.
To learn more about Accrete.AI & IBM visit ibm.biz/AlphaWithAI
See the full 6 minute interview