We ran a hackathon with members of the Programmable Banking Community to see what exciting projects teams could build using the programmable banking tech in a short space of time.
In this demo, Ushik introduces the Investec Wealth Projection platform that his team built to help give users better visibility on their spending habits. The goal is to help users see where their funds are going so that they can better understand how good short-term spending habits lead to long-term benefits and overall improved financial wellbeing. Check it out here!
Check out the repo here.
Transcript of the demo
Ushik: (00:00)
We’ve created the Investec Wealth Projection – we being Pivendren Naik, Ushik, that’s me, and Yurisha. We created this projection tool to project your wealth, and wealth into the future, to improve your spend habits using Azure and Power BI. So our problem was that there’s very little visibility on your spending habits, and how to know the true cost of your spending. If you could see this, the impact could be that you have better spending habits, more understanding of where your funds are going, and how it’s affecting your future. And so the product shows you long-term benefits of short term spending patterns and your overall financial wellbeing.
Ushik: (01:16)
The idea actually came from the fact that knowledge is power, and we wanted to design a platform to represent the true cost of bad spending habits. So, if you spend R10,000 on something that you shouldn’t have, how does it affect you in 30 years time? You don’t usually look down the line to that extent.
Ushik: (01:43)
The inspiration for this came from Dr Calculator – I’m not sure if anyone’s used it, but I’m just going to show it very quickly. I used this quite a lot when I purchased my apartment, so you can set your principal amount, interest and number of years, and you can just play around with different things. So if you’re looking at a 20 year bond, you can actually see in real-time how much interest you are going to be paying versus the actual amount. You can even add things like extra payments. If I add a R100,000 upfront, or how much is this – yeah, I think that’s R100,000 – just how much would that save you in the long run, and how many years does that reduce your bond by? That’s Dr Calculator, and quite a lot of our inspiration for this. In terms of how it works, there’s a post transaction script that allows us to get access to some cash transactions and integration point/scheduler, pulls data of a historic basis, those to merge, and we do reporting on the front end.
Ushik: (03:01)
We built the apps using Blazor and the report was done using LPI. So I’m just going to go through this very quickly because I’m actually going to show this in the demo. Some of the features include your basic registration and login where the user can create an account with their email or they could use Facebook to integrate with the platform. You can also add your phone number, email, password – two factor authentication hasn’t been added yet. That’s something that we are looking to do in future. So once you register you’ll get the script that you copy to your Investec API platform, and this gives us access to your transactions going forward. So this is the report overview that we picked to demonstrate shortly. It’ll show you your income, your overall income for a month, your expenses, and how much you are actually pushing to your savings accounts. ‘m not going to spend too much time on this 'cause I’m actually going to actually demonstrate the entire proper thing. And just some data for where your data – or your funds – actually lie in your Investec platform.
Ushik: (04:28)
So at the top, you can see there’s an overview, there’s transaction history and savings projections in Hong Kong. Your transaction history will show you all your transactions that were saved by the Investec platform. You can select the accounts that you want, and it shows you all of the transactions you had, and it shows you a work cloud based on the amounts that you’ve actually spent. And so this is the savings projections, and based on your different spend types you can actually project that initial cost into 30 years. The home loan projection will give you an estimate of say, if your home cost you R2 million, and how much interest you are actually going to be paying, how much of the funds are actually going to be paying your bond, and what the additional amounts will actually be… So if you spend a R1000 on fast food, how much will that reduce your home loan in the long run.
Ushik: (05:35)
We actually have an invest tab also, which will show you tickers of currently global stocks and the growth projection of some of the funds, like Tesla grew 127%. So if you invested your R24,000 into Tesla a hundred days ago, at the maximum dips and peaks, you were at 30,000 instead of R24,000.
Ushik: (06:12)
Let’s head into the demo. The app, as I mentioned, was built in Blazor, with the registration and account functionality, and this is the script that you are copying and going to the report. So, starting at the overview, as I mentioned, these are all interactive so you can look at costs in your different accounts, and the current balances, your income expenses, and savings, and basically just a mapping of all your costs. Then, if you go to the transaction history you can see how funds have been moving across your accounts, so between, prime linked deposit or prime saver, or things like that. These are all of your transactions. There’s a little work cloud here. So if I click on ‘travel bank account’, it will give you more information on where I actually spend my funds. You can see that quite a few transfers and deposits are fast food. And so if I go down here, select ‘fast food’, you’ll see that all the different fast food pops up. So if I select one of these, it should filter on the transaction that I have.
Ushik: (07:57)
And then the different spenders… So, we can categorise either good or bad spending, and we want to flip this with the machine learning models in the end. Power BI has actually released quite a few machine learning and AI features in their March/April release, so we’re actually going to add quite a bit of that going forward. Savings projections – so, based on your transaction history and the categories that you had, for example, going to select fast food. This is based on one initial deposit, so if you look at this for the track speed investment, your fast food for the previous period was R1,300, at a rate of 3.6%. In 30 years, that’s actually going to be around R4,000. So it’s just an indicator to give you an understanding of where your bad spending goes to, and how much it costs you in the long run, and so a possible return on your investment, based on a single deposit. If you had to invest the funds that make up your income – if you had to invest that into this account for about 3.6%, in 30 years time, you’ll have R118,000.
Ushik: (09:26)
So you can select different percentages to understand how that will grow. What we would like to do with these prime linked deposits, bank saver, and tracks for investments is actually get the interest rate on the day. Currently, I’m not sure – I don’t think that’s available, but yeah. This will just give you an understanding of how to use your money better in the long run. Then you’d be looking at home loan projection, as I mentioned. For example, if your home cost R2 million, average payment based on an interest rate of 7% would be R13,000, your interest will be R2.8 million, and your total payment will be around R5 million, which is quite a lot, and you have your little breakdown here.
Ushik: (10:19)
We haven’t added this in, but we would actually like to do so. So, if you spent R1,300 on fast food, if you click on ‘fast food’, it should show you an estimate of how many years you’re going to be saving the nominal, how much interest you will be saving, and this will give you better insight as to how to spend your money. Then we added in another tab – as I mentioned before – ‘invest’. So currently looking at the global tickers, we are still going to add South African tickers. So in your Standard Bank, Sasol etc to show the growth peaks and how much you can actually save in the long run. Well, this is based on the past 100 days. So if you, as I mentioned before, saved this R24,000, if you invest R24,000 into Sasol, you would actually have R30,000 at the moment. If you invest that into Microsoft, that’d be R5,000. That doesn’t look right. Yeah, that doesn’t look right. I’ll have to look at that.
Ushik: (11:41)
That’s still giving you an indication of how things would actually look in the long run if you saved in different areas. It’s coming soon! Going back to the presentation. Okay, challenges, I’ve locked myself out quite a few times. Pivendren has helped me quite a bit with that. So the issue was that I was using an old key while trying to update and refresh the rest of my data, so that locked me out. Demo data, that’ll be really great to have for testing out some of our accounts or transactions just to get a whole bunch of different data into different Investec portfolios and see that in a way that you can actually build platforms from them. Interest rates, currently not available in an API kind of format, where we can build them on a regular basis.
Ushik: (12:49)
Next steps, ML and AI to detect good events and habits. So if you would like to learn from your past behavior, often you start tagging things, and then it can learn that if I’m marking Nandos as bad behavior, it’ll start tagging KFC or McDonald’s. Also, instead of putting a hard limit and saying Johannesburg is allowed for transactions, you can identify that, okay, this guy is moving closer to Durban on this day, so naturally start allowing transactions in Durban.
Ushik: (13:31)
Then the OpenAPI background sync. So, the merging of transactions needs to be completed. A Facebook messenger bot with notifications – it’s kind of like a little Tamagotchi to show you a little happy face or sad face whenever you make a transaction, and yeah, security. That’s about it. So this is the team, if anyone would like to get in touch and that’s it.
Ben: (14:00)
That’s really epic, you got a whole platform going there, that’s so awesome. How long did it take you guys to build all of that?
Ushik: (14:14)
I’d say around a week, a week of after hour time.
Ben: (14:21)
Okay, I was going to ask.
Renen: (14:24)
Okay, you said it was built in Blazor?
Ushik: (14:28)
Yes, I think Pivendren used a template for the app.
Ben: (14:55)
I was trying to play around with it now. I need to sign into Power BI though. Is that correct?
Ushik: (15:01)
Yes, that’s actually a shortcoming that we saw from the last demo because with Power BI you kind of need to be within the same organisation, but most of this can be transferred to another reporting app. I’m actually not sure when Pivendren is going to fix that, I’ve seen a lot of issues with that.
Nick: (15:25)
Can you walk us through, why did you guys choose the Power BI, rather than say Tableau, or something like that. Is there a specific reason?
Ushik: (15:36)
The main reason is because I have experience in it.
Nick: (15:38)
Okay, yeah.
Ben: (15:43)
Where do you use it?
Ushik: (15:44)
I use it for work.
Ben: (15:49)
Okay, what kind of things do you do with it at work?
Ushik: (15:51)
We do BI for mining data, so we get edge sensor data and we make reports for clients.
Ben: (16:03)
Okay, cool.
Nick: (16:05)
I really loved the whole psychological nudging that you guys got there, with the word cloud and that sort of stuff. There’s a lot of stuff that you can do to help with that, because as you said, that [Walt] nudging sort of thing, which is very cool. That’s something there, it’s like 22seven, but just smarter essentially, so very cool.
Ben: (16:32)
Yeah, using loss aversion as a tactic is pretty smart, showing people what they could’ve had if they’d done different things. You just have so much education around, you could have done lots of things with your money.
Ushik: (16:48)
Yeah, I think one of the main things that I mentioned is that you see the true cost of something, so how much it will cost you, and your future self. I think it’s great to worry about now and leave the future to a future self. But I mean, in the end, it could be depriving you of something if you’re working towards your home or whatever else it is. And for the home projection, it’s not just money that it costs you, it’s time. So 36 months if you can bring that down to 30 months, that’s six months of payments that you don’t have to have.
Ben: (17:35)
Yeah, I figured out in months – how long are car loans? Well, anyway I figured out from the last four months to go that I could actually just pay my car loan in one go, which sounds obvious in retrospect. But time, I was like, oh.
Renen: (17:53)
R2 million bucks for a house – we should all move to Joburg.
Ushik: (17:57)
No, not a house, an apartment – a one bedroom apartment. So I’m not sure if you have used Dr Calculator before, but I really love this and I think it might help anyone who’s looking at homes.
Ben: (18:17)
It’s always great to learn new tools. What was the most fun part for you, or what did you learn the most while doing this?
Ushik: (18:29)
The most fun part was actually getting all of my information together. So, at some stage I started to narrow down that I eat quite a lot of takeout and that’s actually not the best. I didn’t realise it but we used to buy takeout 4 or 5 days out of the week. And I’m actually just getting the little [] up and running, and to see where your costs are coming from is interesting.It helped me alter my behavior on this also.
Ben: (19:04)
Yeah, visibility is massive.
Ushik: (19:08)
Yeah, just being able to see things, see where it’s going and how it’s affecting you.
Nick: (19:24)
I also like the whole integration of using a bot. It is actually very clever, so whenever the transaction gets it – that’s actually very cool too.
Ushik: (19:35)
Yeah, so it’s done really well and also the little Tamagotchi helps with the behavior. So maybe it doesn’t give you a hard limit, because sometimes you do need to make those transactions, but a happy face, a little sad face, a little heart showing your growth – if you put money into your savings account from then the heart fills up. Yeah, things like that.
Nick: (20:08)
Yeah, definitely.
Ben: (20:13)
So Ushik, I’ve tried to set it up on my side by logging into your app, but now I seem to have gotten stuck on this having to sign into Power BI. Is there a workaround for that?
Ushik: (20:25)
I actually don’t think there’s a workaround for that at the moment. But if you have the Power BI desktop, at the moment, I shared the desktop file to the repo. You can basically download that, put in your secret and key and you would be able to do the same and see all the reports.
Ben: (20:53)
Okay, great, because I’d like to play around with it. Cool, any other questions from the floor? No, cool, thanks guys.
Ushik: (21:08)
Thanks guys.
Get involved in the Programmable Banking Community
If you have questions or just want to say hi to the Programmable Banking Community leaders, you can pop us a mail and either Ben or Nick will get back to you.
If you want to see more from what the community has been up to, you can:
