Now Reading
How This Bangalore-Based Startup Is Helping Kirana Stores Harness AI

How This Bangalore-Based Startup Is Helping Kirana Stores Harness AI

Ambika Choudhury
W3Schools

In food and grocery ecosystem, both supply and demand are highly fragmented. It doesn’t help that there is very little transparency on demand and pricing which eventually leads to huge inefficiencies. Even when demand is known to some extent, modern and efficient supply chain infrastructure and reliable payment mechanisms are lacking, resulting in delays, wastage and lost opportunities for producers and buyers throughout the ecosystem. 

With a passion for solving the food and grocery problem in India, Bangalore-based Jumbotail utilising AI and machine learning to fulfill the vision of organising the food and grocery ecosystem in India using technology, data science, and design. Founded by Karthik Venkateswaran and Ashish Jhina in 2015, Jumbotail a two-sided e-commerce platform backed by strong supply chain and payment capabilities. 

It provides an indispensable service to the kirana stores across India. Currently, it has two fulfilment centre complexes outside Bengaluru and eight distribution centres. The supply chain technologies have been completely built inhouse, and the operational processes have been developed to run an efficient supply chain. Jumbotail is leveraging its strong technology, data, and design DNA to build in-house technologies that enable Kirana store owners with limited tech exposure to experience the power of the Jumbotail platform in their native language.



Use of AI and ML

Jumbotail generates billions of data points from a plethora of data sources ranging from multiple mobile apps for customers, sellers, supply chain operations heroes, and other partners from the marketplace platform, supply chain, fintech, and retail customer points of sale. 

The data is widely diversified ranging from transactional data from production systems, application server logs, logistics vehicles/trucks location data, streams of image & sensory data coming from multiple workflows, natural language search text, and third party data from partner ecosystem such as commodity pricing data.

With the help of machine learning methodologies, the company is solving complex problems in wholesale food and grocery commerce. The system drives real-time decision making using machine learning and AI that forms the backbone of the supply chain optimisation systems, recommendation engines, pricing and merchandising systems, credit scoring, and fraud detection.   

Depending on the context and the complexity of the problems, the machine learning developers at Jumbotail use different languages like Python, R, Java, Scala, and a multitude of algorithms which are both proprietary and open source in order to solve the problems. For data visualisation, the company uses Periscope Data and Tableau. According to the founders, the company is building next-gen AI systems using computer vision and deep learning to solve complex problems in supply chain optimisation and retail analytics in highly unstructured category food and grocery.

Core Tech Stack

According to the founders, the tech stack such as choice of languages, technology and methodology constantly keep evolving with the time. The company uses micro service architecture for backend and primarily most of the services are developed in Java using Dropwizard, SpringBoot and Apache Camel frameworks. The developers at Jumbotail use DynamoDB as a NoSql database and RDS as a relational database. 

See Also
IIT-Delhi-Startup-Creates-Buddhi-Kit-To-Make-AI-Learning-A-Child’s-Play_Photo-by-Hitesh-Choudhary-on-Unsplash

Several open-source technologies such as Elastic Search, Redis, Kafka, ZooKeeper, Apache Spark, Debezium and many of the Amazon web services like Elastic compute cloud for Infrastructure, Kinesis streams, Cloudwatch, Redshift as Data Warehouse, S3, Lambda for Serverless  Compute, among others have also been used. 

The company has multiple UI portals used by Internal and External users which are developed in Javascript using React and Angular frameworks and also multiple Native Android apps. For AI and Computer Vision projects, the company use TensorFlow and they also have a full-fledged CI/CD pipeline to automate deployments using Jenkins and Ansible.

Hiring Phase

Currently, the company has a total of 25 members in tech, product, design, and decision science where 15 members belong to the core engineering team. While hiring a candidate, the company looks for someone who has strong fundamentals with an intuitive understanding of core concepts in software engineering and good problem-solving skills.  

Future Roadmap

According to the founders, in the coming years, Jumbotail will expand the J24 stores’ network and help the entrepreneur to transform old Kirana store into modern convenience stores using J24 consumer brand, Jumbotail’s PoS, AI-driven merchandising, omnichannel integrations and a fully integrated supply chain. 

What Do You Think?

If you loved this story, do join our Telegram Community.


Also, you can write for us and be one of the 500+ experts who have contributed stories at AIM. Share your nominations here.

Copyright Analytics India Magazine Pvt Ltd

Scroll To Top