Betting Prediction Software Development Agency

Microservices Migration: 80% Performance Boost for a Prediction Trading Platform

Re-architecting the Scalability & Achieving 80% Performance Growth

Project Highlights

MVP re-engineering for product stability and further scalability

Migrating from monolithic to the microservices architecture

Making a solid solution with poorly pre-selected development approaches

Industry: Fintech

Location: USA

Tech Stack:

PHP
Laravel
Redis
Vue.js
Node.js
Golang
Kafka

Offered Services: Architecture Consulting, Solution Architecture, Backend & Frontend Development, PM

Timeline: 3 months

Platforms: Web

See it live

Client overview

The Project

Douglas Campbell is a professor of economics who has earned multiple honors and industry recognition for a substantial contribution to domain development.

‍

Insight Prediction is a unique trading platform that specializes in betting on markets with single or various outcomes. It allows random people to vote for a certain expected result in a globally important issue with their money.

How It Works

The very same stock trading principle is applied. Just one exception - the trade subject is not stocks or commodities but insight predictions concerning global issues.

‍

When users enter a certain market, they can see a central question and just two or a variety of options to choose from. Those are like single-outcome and multiple-outcome markets. If users believe in the feasibility of a specific outcome, they are able to buy a certain amount of available stocks at a market price and sell it once they expect the price to grow. So they can get a margin from speculations on hype matters. A simple example is a market concentrated on a question of how a football game will turn out for each team - a win, a loss, or a draw.

‍

The platform itself generates these markets and, thus, raises questions that bother humankind. The commission fee is pre-determined by organizers for each particular case.

‍

Challenges To Overcome

We have received an inquiry about the project re-engineering when the admin panel and trading logic were already built by the other development team. The existing minimum viable product was in a poor condition.

The client has decided to switch the development team due to the poor stability of the product and the slow speed of app development.Basically, it was a rough MVP with lots of infrastructures, frontend, and other technical issues.
‍
The project was deployed on one server making the product non-scalable.After a thorough analysis we have outlined the following serious challenges for our experts:

To figure out the business logic

The project was badly documented and fragmented. There was no single source of knowledge so it was hard to figure out how each feature works without technical requirements.

Dealing with poorly selected development approaches

JQuery and Vue.js were introduced for frontend development by the previous development team. Consequently, the frontend part worked badly and it was almost non-fixable. It required us to make occasional remedies as a temporary measure and further total code rewriting.

The transition from monolith to microservices

Monolithic architecture in product development involves the encapsulation of various bots, parcels, third-party integrations, and other services under one solid solution. It leads to the codebase growth and issues with the code maintenance additionally burdened by poor code quality.

Core solutions

Product Optimization

Within the first stage, we have provided services to optimize and stabilize the product's functioning. After just 1-2 weeks from the start of our work, we have introduced the product to a wide circle of people for conducting tests and getting early customer reviews.

Third-Party Services Integration

We have decided to integrate third-party platforms as fast as possible. Literally, we have chosen the path of not only fixing the existing issues but also scaling the functionality to attract more users to the product right away.

Microservices Architecture Implementation

To fix the scalability challenge, we have initiated shifting from monolithic to microservices architecture. This approach allows different parts of code to function separately and be responsible for their own area. It gives flexibility, and adaptability and does not overload the main project. This part included the migration to the cloud services to provide higher stability of the product functionality and scale easily with the needed amount of users.

Frontend Development

The existing solution was created using non-suitable technologies making the product poorly functioning. So we have decided to develop frontend from the ground.

Applied Technologies

The initial technology stack included PHP language, Laravel framework, MySql relational database, Redis for cache, Vue.js, and jQuery for frontend development.

‍

We plan to further utilize extensively innovative technologies like the Node.js framework and the Go language. Occasionally, we may use Kafka for providing communication between bots that parse the information from different sources. Also, the service bot on WebSocket will allow us to update data in real-time.

The outcome

Our development team continues the prediction betting platform development. However, the product has already become stable and faster by 80% within just 1-2 weeks of product optimization.Our expertise in implementing digital solutions within the betting domain allowed us to quickly create a stable product of higher security level.

‍

Even despite the fact that initially it was developed with non-suitable technologies.What's more pleasing is that our client and his investors are completely satisfied with the outcome and are ready to continue developing the product further.

‍

What's Next?

‍

We are planning to expand the platform functionality by integrating it with similar betting platforms.

‍

Thus, the insight prediction product will be able to parse the information from other betting applications and reflect it as a market where the coefficients will be transformed into shares via a special formula. This way, the integrations will scale the betting product even more.