Is there a code in No-code?
Hello! I am Maria Strogonova, I have been implementing custom and standard solutions based on the ProgramBank.BusinessAnalysis storage since 2006. At ProgramBank, I have worked my way up from an analyst to the head of the production center.
The main specialization of our team is automation of business processes data collection and preparation of management and financial reporting. From its inception, the BusinessAnalysis repository was a low-code platform for developers, on which we ourselves automated the tasks of our clients. Therefore, its new version with no-code tools for solving the same problems.
Our user is a planning department employee, a management reporting specialist, a financial analyst, a data analyst, someone who is proficient in Excel, but not a programmer.
Self-service The tools give him the ability to customize his own multi-user applications from scratch, in which
the collection of external primary data into a single storage is configured, this can be downloading from external systems, manual input and downloading from files
Multidimensional reports can be customized
business processes for preparing source data or calculating reports with control over deadlines and division of responsibility can be set up
you can quickly make a copy of the application and make changes or test a hypothesis
separation of access to data and algorithms can be configured
The only task that will require the help of the customer’s IT specialists sooner or later is setting up ETL.
When I talk about our project, I often encounter skepticism in the style: “Well, yes, we saw one like that no-code! This only works for typical sites, mobile applications or hypothesis testing! We'll set up the minimum, and then we'll get hooked on your classic modifications”.
I understand these concerns very well; we use the same arguments, for example, when choosing an import-substituting Service Desk.
But we still want to show that now, even in the area of finance and reporting, it is possible to set up and support a no-code application without programming.
For this purpose, I have prepared a series of articles with specific examples.
My first example is this Setting up budgeting in a small commercial bank.
The article will be of interest to employees of planning departments, management reporting specialists, data analysts, architects and other IT specialists from business who are looking for the best tool for their users or have already chosen our or a similar no-code platform.
Budgeting is generally a classic task for automation on spreadsheets, where ideally all of this is needed – multidimensionality, multi-user mode, business processes, separation of rights, the ability to quickly reconfigure the model for each new period, all sorts of coloring and comparisons.
In order not to delve too deeply into the budgetary jungle, we will take only a small piece – lending to individuals – and using this example we will look at what exactly the user needs to do to set up the budget process in our self-service, and then compare it with the classic approach to automation.
Application setup is usually performed either by the responsible user or according to his methodology, which is drawn up as an excel file. Therefore, during the setup process, we do not ask questions about how many and what analytics or indicators we should have, what is calculated by what formula in the forms – we have all this information.
The structure of the application in self-service objects will be as follows:
Here will be:
3 reference books,
showcase for collecting primary data on the credit portfolio,
5 multidimensional forms.
Let's move on to the direct configuration of the Budget application:
1. Create an application
The new application immediately takes its place in the general application tree.
2. Setting up standard directories
The so-called standard reference books will be further used as financial analytical dimensions in the multidimensional model.
In our example, the standard story for financial reporting is currency reference analysts, financial responsibility centers (type of divisions), products.
Setting up a directory is somewhat similar to setting up database tables, but in terms that are understandable to the user.
The user specifies the name of the directory and configures the list of attributes
names,
their types are string, number, logical or other reference,
is it mandatory?
uniqueness indicator (key).
Attributes for the validity period and settings for unlimited levels and unbalanced hierarchies are added automatically.
The first filling with data can be done by loading from Excel (the standard template is generated automatically), then you can edit it manually directly in the table.
3. Setting up specialized directories
These are service directories where the attributes are already predefined and all that remains is to create them and fill them in.
This is, for example, a calendar, in the settings of which we indicate that the main regularity of multidimensional forms will be a year, and the internal discreteness will be months.
These are also value types – they will be needed to set up different calculation methods and display them separately in the Plan and Actual form.
4. Set up the basic structure of multidimensional indicator forms
We analyze the structure and options for filling in indicators of multidimensional forms in an excel file from a methodologist. Below is an example of how such a description might look.
In the description we see that the form has
analytics parameters: Banking product, Financial responsibility center, Currency,
the columns show the months within the financial year,
by lines – list of indicators,
cells at the intersection – numeric data. Yellow cells – manual input, the rest – calculation by formula.
For the same form in the methodology there are three options – planned values, actual values and Plan-Actual comparison.
What self-service capabilities are there so that the user can customize such forms?
setting up an unbalanced hierarchy of indicators that will be the headings of the table rows, indicating the data type,
selection of analytics (directories) as parameters (filters) of the form,
selection of analytics (directories) as dimensions by rows or columns,
adding value types to indicators.
Already at this stage, we receive a multidimensional form with the required structure and the ability to enter data manually, until the sources and formulas are defined and configured.
We immediately set up aggregation methods for the indicators. It is possible to select standard options – sum, weighted average, quantity, etc., it is possible to set your own formula. These settings are used for automatic calculation of totals and aggregates.
If there is a need to do the so-called vintage analysis in the forms, there is an option to specify the calendar both by rows and by columns at the same time:
Next, you need to move on to setting formulas, so data sources – showcases – should appear.
5. Setting up the data collection showcase
Setting up a data mart is similar to setting up a standard directory
names of indicators,
their types are string, number, logical or other,
is it mandatory?
uniqueness indicator (key),
for numerical indicators, you can also specify an algorithm for calculating the results – choosing from the standard ones: sum, weighted average, etc.
Immediately after saving, the user has access to two complementary ways of filling the showcase
A ready-made Excel template for download can be downloaded from the showcase editor.
At this stage, we immediately add a small amount of data to the showcase for initial debugging of formulas. Later, using an excel template, it will be possible to load working volumes of primary data.
In the future, the bank's IT specialists will be able to automate data loading in the following way:
through the Business Dictionary setup – then the user will be able to set up the connection of the showcase with an external system, choosing ready-made indicators among the names of sources and their attributes that are understandable to them
by setting up data loading from an external system directly into the showcase table
6. Setting up formulas
Now comes the most time-consuming part of the process of setting up multidimensional forms – formulas.
In formulas, the user can use the reference to
shop window indicators
other indicators within the same form
indicators of other forms of the same application
indicators of forms of other applications
indicators with a different context – from other periods and by other analysts
aggregated indicators
In addition to basic operations of addition, multiplication, etc., the user has access to analogues of Excel functions.
There is a function navigator from which you can add call templates.
There is a navigator for application indicators, from which you can quickly add the rest of the formula content.
What functions, for example, can be used in formulas
Similar names to Excel
SUMMESL
VPR
OSPLT
PRPLT
SUMBELOW – the sum of all child indicators in the hierarchy
AVERAGEWEIGHTIF is an analogue of SUMIF, only the values are not summed up, but the weighted average is determined
When accessing from one form to another, the context itself is “glued together”, and the missed analytics are aggregated themselves, taking into account the previously made settings.
That is, in principle, we strive for minimalist formulas and many obvious things are performed automatically.
Examples of formulas:
"Выдано кредитов, тыс.руб" - СуммЕсли("Кредитный портфель#Сумма повторных выдач по траншам ВКЛ, руб")/1000
here the data of the showcase for summation is automatically filtered by the product from the context of the cell with the formula
"Возврат кредитов, тыс.руб"* "стадия обесценения 1, тыс.руб "[[Предыдущий "Год"]]/"Портфель на конец периода, тыс.руб"[[Предыдущий "Год"]]
calculation based on previous year's values
"Портфель аннуитетных кредитов#Начисление процентов, тыс.руб"[["Банковские продукты" = 'Ипотека ФЛ']]
we take a value for a specific product from another form, while another context or value type is not specified, which means they are automatically determined as matching the call cell
СрВзвешЕсли( "Кредитный портфель#Срок кредита (транша), дней", "Кредитный портфель#Сумма выдач, руб", "Кредитный портфель#Сумма выдач, ед.">0")
For formula functions such as SumIf, AvgWeightIf, users receive automatic detailing of the values calculated for them.
Also, using the familiar formula language, users can set up conditional formatting by indicators, for example, to emphasize critical or, on the contrary, super-nice values.
Bonus with showcases for external access to collected data:
Users with sufficient rights can connect directly to the storage via sql and use the collected and calculated data right away in their favorite external tools like BI or Power Query. For this purpose, each showcase and multidimensional form has auto-generated views.
7. Setting up a budgeting business process
So, the sources, structures and formulas for the forms are set up – now it's time for the business process.
Here too, the visual setup is quite simple.
Need to determine
names of steps and stages
their sequence and dependencies
list of responsible users
a form that needs to be filled out at each step
All.
When launching a business process for a specific financial period – in our case, the next planning year – the process administrator will specify relative or absolute deadlines for each step.
In accordance with this specific regulation, each user will be automatically notified of upcoming and expiring events, and will also see a list of their current tasks each time they log into the self-service application portal.
8. The final structure of the self-service application
As a result, our financial analyst received the following functionality
business process planning
loan portfolio showcase with 4 complementary data collection options
multidimensional forms in terms of currencies, financial responsibility centers and products, as well as their aggregates:
Vintage Credit Analysis Matrix
Portfolio of annuity loans
Model of reserve formation
Financial performance report
Balance
independent further management of changes within the bank
The final output of this application will be Balance Sheets and Financial Performance Reports for each financial responsibility center and consolidated reports for the bank as a whole.
What was technically done for the setup:
Entering names
enabling options via checkboxes
selecting values from a list of options with common, non-specialized names
Organizing lists with the mouse
writing formulas similar to those in Excel
Of course, the example described is the automation of only a piece of the bank's budget; some of the indicators in this example remained for manual input.
Nevertheless, some conclusions can already be drawn.
It took a trained analyst up to 8 hours of pure time to set up this functionality on our self-service platform. Let's say with breaks for thinking or consultations, it could have been ∓3 working days.
What do we know about what he could have spent the same amount of time on?
Yes, only for writing technical specifications for automation of the same thing.
Compared to classical automation, when using a self-service platform, the bank receives
saving employees timesince they will be able to implement applications within the timeframe in which the technical specifications are usually written
reduction of operating costssince the number of applications on one platform is not limited
there is no need for employees with any special qualifications to work with the platform and further support applications
In our opinion, honest No-Code looks exactly like this.
Please provide feedback in the comments on what points you found interesting and what would be useful to explore further in the following examples.
I am also interested in your opinion on whether it is correct to call such instruments no-code or not.
If you are interested in the BusinessAnalysis platform, we suggest you watch the video recording of the webinar “How to create and configure multidimensional budget model of a commercial bank without MS Excel and programming” or attend our new webinar in September'24.
At webinars, we not only talk in more detail about the capabilities of our self-service tools, but also literally show everything that I wrote about above – namely, setting up applications from scratch.
