About Ntropy
Ntropy is the most powerful multi-geo, multi-lingual transaction enrichment API. Technical briefing by Ilia Zintchenko, Robin Kahlow, Naré Vardanyan, and Michael Jenkins.
Overview of Briefing
0:00-1:11
Ilia Zintchenko, Co-Founder and CTO, opens up the demo with a quick background about the problem their tech is solving. Ntropy’s vision is to enable companies to understand their users through their bank transaction histories and other financial data.
1:12-2:02
Ilia starts out by going through the onboarding process. We can see it is easy to login using Google’s SSO option. Basic KYC information is collected during this process as well.
From here, we see the dashboard for Ntropy’s application. Ilia then quickly creates an API key with the click of a button.
2:02-4:22
After Ilia copies the API key, we move to his computer Terminal to paste. Ilia uses Ntropy’s docs to grab a use case this API key has. Ilia copies and pastes some code from the docs into his terminal where we see it only takes a few lines of code to enrich transactions.
Ilia quickly generates 12-18 months of transaction data to use in the demo. Once everything is imputed where it needs to go, it takes just a few seconds for the enriched transaction file to be ready.
Ilia takes a look at the enriched file and runs through a couple examples of what data is collected from these transactions.
4:52-7:27
Ilia jumps back to Ntropy’s application to view this data in a friendlier way. He quickly uploads the CSV file that was generated using the API key earlier in the demo. After a few seconds, we are shown a friendly view of the transactions. Here we can see details about all the transactions and any recurring payments.
We also see income sources, which shows how often income comes into an account and a yearly total. Ilia shows what the API pulls together with the information given from the initial transaction file.
7:28-8:38
Next, Ilia moves into answering a hypothetical question, can ChatGPT complete this function?
Ilia explains that Ntropy is ~2000x cheaper, ~150x faster, and ~15% more accurate than ChatGPT.
8:39-10:35
Ilia passes off the rest of the demo to Robin Kahlow, machine learning engineer at Ntropy. Robin is going to present a fun use case using Ntropy’s API.
Robin starts by giving some background on transaction monitoring. We learn that current banking applications have set features, with little to no optimization for the end user. One solution to this, using Ntropy’s API, is Ntropy Cookie, a chatbot for your transactions.
Robin quickly gives an overview of this use case before diving into showing how this can be done easily using Ntropy’s API.
10:36-12:38
Robin dives into demoing Ntropy Cookie, mentioning this is not a part of any product that is currently offered.
Robin uploads the same transaction history that was used earlier in the demo into the Discord chatbot. After completion, Robin shows that you can ask the chatbot anything about the transactions. We see an example where we ask how much was spent on fast food, to which the chatbot responds with a summed up list of transactions at fast food restaurants.
Next, Robin shows some pre-existing prompts. We use the “essentials” prompt, which breaks down between essential and nonessential spends. Robin corrects the chatbot about some transactions that were listed in the incorrect group by simply writing back. The groups were fixed immediately by the chatbot.
12:39-13:51
Another prompt Robin shows us is “saving opportunities.” The chatbot lists out repeated transactions that are considered “non-essential” and opportunities the user can take to increase the amount of money that can be saved.
If you are interested in trying out this feature, reach out to the Ntropy team!
Additional Briefing Information
This briefing was held at V-Sum_Twenty on May 16th, 2023. V-Sum_Twenty was made possible by IowaEDA and Brale