Spoiler alert! The answer is yes. No matter what size is your business and if it has Data Scientists or Data Analysts, it can start the ML journey right away. In this blog post I want to go over several examples of use cases that we are working on around the world and how CTOs from small to large businesses are implementing Machine Learning without having to spend big resources both in terms of personnel or monetary. 


Recently I shared a tweet where I stated how new technologies are speeding up the ML adoption and how this is augmenting human intelligence to impact in a positive way how companies operate and hence improve their results. That said, the same way technologies like Repl.it are helping people learn how to code without the need of powerful computers, technologies like Datagran are helping companies that previously believed they were not ready, to adopt ML, to implement it and get the results they are looking for.


The first thing I would like to point out is that this responsibility lies over the shoulders of CTO’s or IT departments. Across the board we see an increasing pressure on these professionals to adopt data architectures that are advanced and that can feed the different teams within companies. For that reason, CTO’s are having to make hard decisions in terms of how they balance the implementation of product and historically complex implementations of Data pipelines and everything that involves building and putting ML models into production.


The first example comes from the CTO of a multinational company. This company is in the Quick Service Restaurant industry. As most companies, this one in particular has a focus in data collection and analysis to be able to understand their customers and hence serve them much better. Recently the company has put together a Data Science team but the team is relatively new and not only that, there are no data engineers. The CTO faces a huge dilemma. One option is to build from scratch data pipelines, for that he will need to quickly hire data engineers and he will need to allocate several resources to the task which would mean to deviate from other important responsibilities that have a higher priority. This is actually causing major delays in other organizations. One of them told me that they are waiting for a churn model they asked for 6 months ago and that it keeps getting pushed out because other priorities keep arising. Another option would be to adopt a tool like Datagran that can provide enterprise level security and reliability to automate data pipelines for their newly created data teams and help them quickly deliver results. 


In this case the CTO decided to adopt datagran.io to help the data team deliver and in the meantime he could understand how the team and the organization want to take the ML journey. Will they want to create a Core team that serves the organization and then embed Data Scientists in the Business units? Will the organization only focus on e-commerce for now?

Either way, implementing a solution that’s quick and reliable at the same time gives the organization what they need when they need it but not only that, it gives the organization a chance to learn and adapt as they go.


This decision of the CTO generated a snowball effect across the whole organization, making teams in different countries adopt Datagran and quickly show business results which can then fuel the CTO with arguments to build what his teams need according to the experience generated by the different experiments.



The other example comes from a small startup that helps independent retailers to buy goods from larger distributors. This startup has an app where small retailers can choose the products they will buy every month. One of the things that retailers could benefit from is an intelligent system where the app could recommend what products should the retailer buy depending on past purchases and what other similar retailers are purchasing.


The problem is that the Startup doesn’t have either a Data Scientist nor data engineers. Usually these startups would be left to one option which is to grow and later implement features like this one that could benefit their clients. NOT ANYMORE. 

The CTO of the startup decided to hire Datagran to not only implement their data pipelines but to also provide a solutions architect to build the models that would recommend the products to retailers. All of it was implemented in a matter of days without deviating the CTO and team from their core product.


Conclusion:


Teams big and small can now implement Machine Learning to improve product capabilities without having to spend large amounts of money and resources. Previously large teams including Data Engineers, Developers and Data Scientists would be needed to implement such projects. Nowadays a few weeks are needed to hugely impact the product capabilities and how companies serve their customers.