With its ride-hailing, last-mile deliveries, and financial services solutions, Yassir app is an amazing software that helps users in over 45 cities in Algeria, Morocco, Tunisia, South Africa, and Senegal with their everyday lives. These users are both customers and suppliers, such as couriers, drivers, eateries, and others who utilise their platform to manage their operations.
It process a wide range of information at Yassir to make sure it give users the finest and most dependable solutions possible across all of it products, and it rely on that data to keep improving those services. Data and AI were hard to integrate with old infrastructure.
It previously utilised Google Cloud and BigQuery for data storage and analysis and Databricks for machine learning model deployment and training. It couldn’t address formatting inconsistencies caused by this setting. Furthermore, it was not feasible to retrieve data from Databricks for processing within Google Cloud, which had a direct effect on the speed of their application.
Despite paying to maintain distinct environments and frequently having to duplicate effort to build and manage any data initiatives, Yassir app teams were unable to obtain the information they need at the right pace because of these fragmented environments.
It made the decision to use Google Cloud to unify it data infrastructure in order to solve these problems and centralise all of these operations. Better data access and scalability would be made possible by this shift, which would also open up new avenues for performance analysis, evaluation, and improvement.
Creating a more flexible, unified data platform
Yassir current partnership with the Google Cloud team gave us a solid basis on which to build new data processing workflows with BigQuery, deploy new Artificial Intelligence and machine learning models with Vertex AI, and overcome the data connection issues. It were able to evaluate and manage costs centrally and implement straightforward, centralised data governance rules by consolidating with a single data supplier. It can test and iterate without making a firm commitment to every project since, as a growing business, It can scale the cloud usage up or down to optimise expenses. This flexibility is priceless.
To create a solution that supports expansion objectives, Yassir app collaborated closely with the Google Cloud team. To assist their team learn the ins and outs of BigQuery and its real-time, governance, and open-source features, It has to take part in technical and strategic seminars. This gave its engineers the tools and resources they needed to explore. By working together, it can foster the kind of engineering culture hope to establish at Yassir app. Rather than relying solely on unconventional solutions, it can address more complicated issues by tailoring adaptable, pre-existing technologies to unique use cases.
Yassir app moved individual models from their old solution into Vertex AI to evaluate their consistency after completing it was internal compatibility checks, and they are now operating almost independently. It has increased the effectiveness and efficiency of machine learning procedures and put themselves in a better position for future expansion by switching from Databricks to BigQuery and integrating it’s own models with those offered by Google Cloud. Even while it don’t now process petabytes of data, Yassir is aware that it can in the future.
Evolving from data processing to data insights
It was challenging for us to give certain teams secure access to particular data because of it previously disjointed data solutions. handing a person or team access to information means handing them the keys to everything because it used Databricks to deploy models but kept the data in BigQuery. Role-based access restrictions and Infrastructure as Code (IaC) Terraform scripts may now be used to automatically give and revoke access to datasets for teams or individuals. It can also guarantee that the necessary data reaches the appropriate users by sharing data via Looker Studio Pro and giving more technical users direct access to BigQuery tables.
Yassir app can better assist everything from customer growth and retention to marketplace optimisation by offering insights into product usage, customer data, and more with data being unified in BigQuery and linked to machine learning models. It keep a careful eye on operational and analytical statistics and develop dashboards to make sure it is meeting their internal and customer-related objectives.
Yassir app sales and marketing teams can better target and contact merchants and customers with the help of it’s operational dashboards. They also provide information on hiring procedures, which enables us to enhance the way it serve certain markets, finish more trips more quickly, and progressively shorten delivery times. In order to detect fraudulent visits and orders and to enable real-time dynamic pricing, it also have product-level detection and monitoring. Yassir app has more chances to create a more individualised and reliable client experience with every piece of data were gather.
Yassir app leadership team uses quickly accessible datasets to inform strategic choices, such as identifying the business’s most critical areas, making macro-level plans for growth trajectories and marketing budgets, and making regional investment decisions to expand the company. Their whole growth plan is based on these roadmap decisions, which would not be feasible without the flexibility and scalability that BigQuery has allowed us to achieve.
