Happy Wednesday!
I am aware it’s been a while since I last wrote this column - it is not due to lack of interest in showcasing our excellent founders & new ideas, but rather, due to lack of time. Things have been HECTIC - and while I love Startups to Watch, sometimes we need to pick our battles, that’s why I’ve focused more on the LatAm Tech Weekly lately. I will try, nevertheles, to make this post more frequent. I promise :)
For those who are new, this series is dedicated to early stage startups from LatAm that are under the radar – and have a good story to tell. I used to send it out every other Wednesday - and it is separate from my traditional weekly newsletter.
This week, I will share a company that is focused on the topic of the moment, that is literally EVERYWHERE - yes, you nailed it: Artificial Intelliegnce. Read further to get to know more about Hyperplane.
Finally, founders this message is for you: if you want to Be showcased here, please help me out and reach out directly to me following the instructions below.
Please fill out a one-pager, in English, about your company using the topics and themes outlined in the posts. Please write it in 3rd person.
Reply to this e-mail with your one-pager attached.
I will review and reach out if I need more information.
It will be posted if I believe the startup fills the profile and audience of this newsletter.
Good Luck!!!!!!
Note that the basic criteria is a company that is not on the spotlight and is NOT actively raising proceeds. There is also no bias – apart from the fact that I like the story. I am not an investor or get any financial compensation to promote startups featured. This is intended for informational purposes only, using my solid base of subscribers to share intel of good people doing cool things, thus fostering the ecosystem.
Since launch, featured startups in aggregate were connected 111 times, to 51 different investors/strategic partners/companies globally.
Follow me on LinkedIn , Instagram or Twitter for daily updates!
Opinions expressed here are solely my own and does not represent those of people, institutions, organizations that I may or may not be associated with in any capacity, unless explicitly stated.
Context
Despite the widespread adoption of AI in financial services, its full potential still needs to be explored. Historically, banks have utilized traditional machine learning models to process structured data, enhancing decision-making in areas like credit underwriting and fraud detection. However, large language models (LLMs) mark a paradigm shift, enabling the analysis of vast amounts of unstructured data such as customer interactions and transaction details. This capability allows financial institutions to gain deeper insights into their customer’s needs and preferences, heralding one of the most significant transformations in the financial markets in decades.
Consumers frequently manage multiple bank accounts in the digital age, intensifying the competition among incumbents and neobanks to become the primary bank. This status significantly impacts a bank’s financial metrics like cost of capital, transaction volume, and overall lifetime value (LTV). According to a survey by Idwall, a digital identity management company, the average Brazilian holds six bank accounts. Although no industry standard exists for measuring a bank’s primacy, it generally involves three components: the percentage of deposits, spending, and financial products held at a single institution.
By harnessing AI to understand and respond to individual financial moments and preferences, banks can offer hyper-personalized products that create more relevant and engaging experiences. We’re heading into a future where each interaction with our banks is enriching, educational, and personalized. AI will empower banks to put a personal banker in every customer’s pocket, and the results will transform our society.
Problem
Banks need to be ready to be AI First.
Enabling hyper-personalization across product lines in the financial system requires massive research, talent, and infrastructure investments.
In their current state, in most cases, financial institutions only extract a fraction of the value from their first-party data. Data is siloed, ML talent is inexistent or underprepared, and very little research goes into extracting value from unstructured data. Navigating this platform shift within such large organizations is a massive undertaking.
The Solution
Hyperplane is bringing AI to banks. They empower financial institutions to create hyper-personalized consumer experiences through specialized foundation models (LLMs).
Their platform seamlessly ingests first-party data, so FIs can effectively train, evaluate, deploy, and monitor any predictive model powered by a complete picture of their users.
By understanding the users’ intent from unstructured data such as transactions and user events, Hyperplane’s LLMs can capture the user’s financial behavior to power consistent personalization in next best offers, cross-selling, and CRM strategies, as well as credit underwriting, credit limit estimation, fraud, collections monitoring, and many other verticals within the bank.
The Platform
Hyperplane’s platform integrates into customers’ existing data lakes, warehouses, and cloud storage layers. First-party data is then continuously ingested and fed into AI modules that are purpose-built for different business areas’ needs. The modules’ output is consumed via API or directly imported into existing data lakes or warehouses.
With their platform, data science teams seamlessly fine-tune foundation models on downstream tasks to power decisions that drive improvements to growth, profitability, and, ultimately, every aspect of the business.
By streamlining the process of creating, deploying, and monitoring a new model, data science teams can solve 10x more problems using AI and improve the time to production of their AI models.
Want to see the platform in action?
Traction
A large financial bank today is multi-product. Many have more than a hundred customer conversations simultaneously and financial products on different channels.
Understanding what moment in life each customer is going through, the next best offer, and when to improve the “share of wallet” is important.
Hyperplane already serves 10 of Brazil’s most prominent financial institutions, collectively serving over 50% of the population. The company has just been selected to participate in the FinTech Innovation Lab, which is a program founded and run by Accenture and the Partnership Fund for New York City, helping early-stage fintechs accelerate product and business development by working closely with some large financial institutions in the US.
Team
Hyperplane was founded by Felipe Lamounier, Daniel Silva, Rohan Ramanath, and Felipe Meneses. Daniel and Rohan spent the better part of the last decade building large-scale AI systems at Google AI and LinkedIn, where they experienced first-hand how centralized intelligence is leveraged to transform the way companies use their data to model, understand, and interact with their users. Now, at Hyperplane, they are bringing that same stack to the financial industry by enabling banks to use the entirety of their internal data to build a central intelligence layer that powers all of their prediction models and customer experiences.
Hyperplane has raised US$ 6M in investments from Lachy Groom, SV Angel, Clocktower Technology Ventures, Liquid2 Ventures, Crestone VC, Soma Capital, Latitud, and Atman Capital.
Website: www.hyperplane.ai