Meet Chanakya: The Platform Behind Anko’s Data-Driven Solutions

Meet Chanakya: The Platform Behind Anko’s Data-Driven Solutions

Anko's Data Science Dream Team has pulled out all the stops and created a platform that's going to change the game - introducing Chanakya, the sophisticated solution to all our data science woes!

A little background - Kmart Group's Data Science team builds AI-enabled data products which have a well-defined lifecycle: data sourcing, cleansing, exploratory data analysis, statistical inference, machine learning algorithms, prediction, and visualisation and activation of findings. 

A significant amount of time was spent on setting up the cloud infrastructure and building the software components for each of the activities listed above leading to repeated effort across all Data Science projects. Can we automate this, make it more efficient so that our team can focus on the right tasks?

With Chanakya, yes, we can! To increase the efficiency of project delivery and allow faster experimentation, the Data Science team at Anko, have built a sophisticated software platform on cloud (AWS) that provides our Data Scientists & Engineers all the building blocks (pre-defined templates & services built using best practices) at various stages of the Data Science lifecycle. Data Scientists can now focus on solving business problems by building great models and let Chanakya worry about infrastructure (computing resources), load balancing, scaling, scheduling, monitoring, etc.

Some of the components that we have built from ground-up are:

  • Data & Analytics Kubernetes Cluster on Amazon's EKS to host all our applications
  • Centralised & auto-scaled Development Environment for Data Science (JupyterHub)
  • Centralised & auto-scaled Workflow Management using Airflow for Data and ML pipeline
  • Custom Python SDKs (Software Development Kits) with high re-usability
  • Simplified Monitoring, Logging & Alerting.

 

This puts us in a select group of Retailers in the world that have an end-to-end in-house platform enabling better data driven decision making at low cost. 

With over 30 active users, including data engineers, data scientists, and business users, Chanakya has already made a huge impact, supporting over 100 data pipelines and 25 machine learning pipelines in production. And it's not just a pretty face - Chanakya has helped save big bucks, with an efficiency increase of 20-30% and a significant reduction in execution time for data science projects.

In conclusion, Chanakya is the missing piece in the data science puzzle. It's the platform that helps build data products with ease, eliminating redundant tasks and saving time and money. Its absence causes duplication of effort, increased time, and increased expense. Data scientists use this platform to create solutions with a streamlined deployment process that does away with the need for people to spend time configuring infrastructure and software components. For Anko, building data products and servicing customers has never been simpler than with Chanakya!


Written by: Kunal Ramkumar, Principal Data Scientist at Anko

Great work Kunal, as always!

Like
Reply
Pavan Kumar Reddy Sannadi

Software Engineer 2 at Microsoft || Data Engineer || VIT

1y

Kunal Ramkumar n team great going 🙌🏻👏

Like
Reply
Somsubhra Sikdar

Head of Data Science @ Anko | Monetizing AI/ML Solutions

1y

Kudos to the team who has built it completely in-house and with zero cost. This, in the long run will help us scale with significantly lower time to market.

Rajasimha Karanam

Head Of Engineering & Delivery | Technology Leader | RetailTech | FinTech

1y

Cool stuff!!! Great read Kunal... 👍

Like
Reply

To view or add a comment, sign in

Insights from the community

Others also viewed

Explore topics