Gopu Pillai: Shortest time to conclude PoC for Enterprise software

Hui Fang Yeo

Discovering Atoti as a self-service BI tool

Gopu Pillai is a Business Analytics and Insight Lead for the IBOR Transition programme for Standard Chartered Bank.

With the cessation of LIBOR (London Inter-Bank Offered Rate) by the end of December 2021, Gopu and his team faced increasing requests from the various business streams on the IBOR (interbank offered rates) exposure.

As the existing system started straining under the increasing data load, Gopu started evaluating alternative solutions that could support the Bank’s needs. Read on to find out how he discovered Atoti and decided on the BI tool.

Huifang: Could you please share with us about yourself?

Gopu: I’ve been with Standard Chartered Bank for seven to eight years now. Previously I was with the Market Data team. I’m now the Business Analytics and Insights Lead for the IBOR Transition programme. 

What we have been doing essentially is looking at the Bank’s overall IBOR portfolio, collating the exposure from the various business streams for regulatory reporting and overseeing how we can move from IBOR to RFR (Risk-Free Rates).

The Business Analytics and Insight team is divided into Business and Technology sections. I take care of the technology section, looking at what technology we have currently deployed and if they satisfied our business needs. If the needs are not met, I’ll look for alternative solutions that can better serve our purpose and report it to the senior management. 

Huifang: Do you face any challenges?

Gopu: The way we collected the IBOR exposures and reported them back to the regulators have changed over the period of time. 

For the last year, we used to collect the exposure at the end of every quarter, either manually or via a feed. Using dataIQ to perform the ETL, we get the output consisting of a million rows of data with 200 plus columns in CSV format. Then we loaded the exposures into the data accelerator before going into MicroStrategy, a BI tool, for reporting. 

It worked well until late last year. We were only doing it on a monthly basis and we reported back minimal amount of data. The frequency of reporting in the past two quarters has gone from quarterly reporting to monthly.

We needed the historical data to be able to show the trend of the insight. However, with two months worth of data in MicroStrategy, it was showing us the “spinning wheel problem”.

The spinning wheel appears when the query takes too long to respond.
The “spinning wheel” shows up when data takes a long time to load.

Users can easily wait for 5 minutes for the report to load. Every parameter change will render another 5 minutes for it to get reloaded. 

It got to a point where users started shying away from using MicroStrategy. Instead, they started writing emails directly to the Business Analytics team, asking for the data extraction or Excel with some Pivot tables for their usage purpose. 

With the cessation of LIBOR by the end of 2021, we are now facing a lot of requests from various businesses regarding the state of the data, the kind of exposures that we are looking at and what we are reporting back to regulators. We increasingly face these questions repetitively month-on-month. The team is overwhelmed by these requests.

Huifang: What actions do you take to overcome these challenges?

Gopu: Instead of focusing a lot of resources into catering to these requests which was too much for us, we thought we needed to come up with better tools that could also be “self-service”.

We started evaluating the BI tools that the bank is using internally. Other than MicroStrategy, we have Tableau and ActivePivot, which is the Java enterprise version of Atoti. We needed a tool that could replicate some of the functionalities that we have in Excel, especially the pivoting. 

Huifang: What are your considerations when evaluating these tools?

Gopu: Tableau was not taken up due to licensing issues and the sheer cost of implementation, in terms of the time of implementation and time to production.

With regards to ActivePivot, we tried to reach out to our internal team who was using it. However, they were very busy with their own project. Since ActivePivot is a licensed product, I cannot simply install it internally in one of our servers and start evaluation right away. That’s when I started looking up the product on the internet and came across its community version, Atoti.

With the community edition, I could download it on my home computer and start working with it. 

I had two machines-a Mac and a Windows. Atoti worked perfectly on my Mac but I ran into some issues with the installation of the tool on Windows. However, I was able to reach out to the Atoti community for help. 

I spoke to Huifang Yeo from the Atoti community team about the problems that we are trying to resolve in the bank. So, she brought in her colleague, Xavier Pilas to help perform a PoC (proof of concept). 

Huifang: What are the key aspects of the PoC?

Gopu: I explained the key requirements to them:

  • Ability to handle massive data
  • Central platform for people to view the analysis
  • Has pivoting functions similar to that in Excel
  • Entitlements at granular levels

By sending data in Excel to the business, we could have some audit and reporting issues as anyone is free to update the spreadsheet.

In the targeted solution, we want to have a centralized platform where nobody can edit the data. However, users should be able to build their own pivot tables on top of the data, so that they can get all the insights that they require.

It should be a kind of self-service tool that everybody knows how to use.

In addition, we needed entitlements at a very granular level. For instance, given that a trade can be booked anywhere. If the currency is, for example, the Singapore dollar, only a certain set of users are allowed to view this data.

Huifang: How did the PoC go?

Gopu: With the given problem statement, we completed the entire PoC within two weeks end to end using Atoti+. It’s the enterprise version of Atoti. 

The PoC was built on the AWS platform and we accessed the Atoti web application from a given link. We could then showcase to the senior management how the pivoting could be done on 1 million data points.

In fact, we extended the 1 million data points by 24 months so as to have 24 million data points. It is to ensure that we are able to scale up to that extent without facing any issues. 

We performed the PoC on these 3 fronts:

  1. The replacement of Excel
  2. The charts along with the credit analysis
  3. The entitlements

PoC findings

From the PoC, we no longer see the latency in data that we used to have with MicroStrategy. Atoti was very quick in processing the data, maybe because it was using an in-memory data cube.

We saw from Poc that Atoti was very quick in processing data
Build responsive pivot table with interactive content editor

We needed to showcase to the management how as the bank we have been performing, in remediating or minimizing the usage of Libor within the bank. So having the trend analysis in Atoti is useful in showing the trend of exposure for different business streams, from the past year to where the exposure is currently at. The trend analysis was something that we could have done with MicroStrategy, however with some latency. 

During the PoC, we were able to create charts in Atoti, which comes in handy for trends analysis.
Trends using Line chart in Atoti

We can also create new metrics quickly in Atoti. In 5 – 10 lines of Python code, I was able to create different metrics and have that displayed in the UI. Users can immediately use those metrics to see the data.

We take the time taken to implement as a factor in our consideration. You must know that a typical IT project can easily span six months to a year. We want something that can be implemented quickly with minimal effort and is not so resource-intensive.

We did a lot of PoC within the two weeks and showed the senior management that we are able to resolve the issues that we are facing with Atoti. Knowing this, it shouldn’t take longer than two months if we have to implement it in the bank.

It also helps that we can just deploy the Atoti library on the server and load data into the cube. It is able to show us all the metrics in the UI.

Huifang: Has the PoC met or exceeded your expectations?

Gopu: Initially we had no intention to replace MicroStrategy. We were building charts with it and we had our templates. On the other hand, we meant to use Atoti for the pivoting of data so as to replace the usage of Excel. So these were two different use cases.

But during the POC, we saw that Atoti also offers a variety of charts in their web application, which could be made use of. That’s when our thought process went ahead and we started thinking about using Atoti to construct those charts as well. 

Instead of maintaining two different tools, we could eventually move all our business analytics and insights to Atoti.

Huifang: Is it easy for you to kickstart with Atoti?

I wouldn’t say I’m an expert in Python but my bare minimum knowledge of the language was good enough for performing the PoC with Atoti. 

I could install the Atoti packages following a few simple lines of instructions from the Atoti documentation. With the sample code in the tutorial, I was able to spin up a cube and get the UI running.

As a BI tool, Atoti easily spins up a BI analytics platform with a few lines of codes.
Spinning up a web application with a few lines of codes

Huifang: What is the best thing you like about Atoti?

Gopu: Firstly, it’s the ease of use of the tool.

Also, the availability of the community version – I would be more hesitant to try out Atoti if there is no community version. Because then, I had to look at the license issue and things like that. With the availability of the community edition, I could just download it on my home PC and see how the tools work.

From my access to the community team till the time we conclude the PoC in two weeks, it’s something that I couldn’t imagine with any enterprise software. At least it’s unheard of in my experience.

Even though the community edition doesn’t include some of the components such as the entitlements, it is a very handy tool that you can easily download and install. You can just import your data into Atoti and see them on the browser. 

Simply drag and drop the rows and columns that you need for the pivot tables and charts. You can see all the insights you need immediately.

Many thanks to Gopu for the interview and for providing valuable feedback. This helps us understand the needs of our community. 

With the Atoti community edition, we hope everyone will have the freedom to try the tool at their own pace. Experience firsthand the features it offers. Do join the Atoti Gitter community for updates and support.

Like this post ? Please share

Latest Articles

View All

Hierarchy Fundamentals in Atoti

Read More

Retail Banking Analytics with Atoti

Read More

Putting Python Code in Production

Read More


Join our Community for the latest insights and information about Atoti