How I created AI based SaaS product along with my full time job?

Above picture makes it easier for me to describe what I am doing as a side project. Its called Lipishala . Lipishala is a word of Hindi origin, meaning home of languages. My core idea /vision behind Lipishala is to enable Natural Language Understanding ecosystem for Indian Languages in form of easy to integrate APIs hosted on cloud.
Before I delve into the details of evolution of Lipishala let me tell that I am working in full time job in the capacity of Tech Architect. I have been lucky to go through the development of product from its ideation phase to a decent scale. In my day to day job I look after the data pipelines and work flows and lead machine learning initiatives. The ML initiatives that I lead revolve around recommendation engines and fraud detection.
How do the idea of Lipishala came up?
I have been very curious and appreciative of latest products offering services for Indian Semi urban and rural population especially in their native languages like Hindi, Tamil, Bengali etc. That ignited my interest and I started looking into the Eco-system of Indian regional languages more from a machine learning guy’s perspective. My early day analysis told me that there was research work being done regarding similar problems but there is nothing available as simple utility which could help businesses build smart tools for Indian language content consumers.
How much time I dedicate to my side project?

I am a father of 3 year old angel and I need to fulfill my duty of a father, after coming from office. So on a normal day I start my work between 10:30 PM to 11 PM when she sleeps. So for past 6 months I have been trying to be consistent and dedicate 2 -3 hours on minimum basis. On weekends I try to put in 8–10 hours. So on weekly basis I try to clock in between 15–25 hours. Other thing that I do 24x7 is whenever any idea flashes into my mind I try to document it my Google Keep notes. This helps me in making most productive use of my work hours.
What is the status now?
Lipishala proudly boasts of machine learning services for Hindi language today. A basic site detailing the offering and having a demo for Hindi is live. I am discussing the product with friends and letting people know that they can integrate the API’s and creating interesting stuff for Hindi users.
Dr. Partha Pratim Roy from IIT Roorkee, an expert in NLP, Pattern Recognition and Multilingual Text Recognition, has agreed to mentor us in this initiative.
And we are have started work in other Indian languages. Tamil, Telugu and Bengali to be more precise.
That was most of it so far. I will keep posting Lipishala’s journey here and look forward to listening to readers thoughts, feedbacks and advices.
Will like to end this post with one of the inspiring quotes that I keep revising whenever I am down.
“The gamble isn’t on our dream. The gamble is on us.”
Good bye folks. Thanks for reading.

This story is published in The Startup, Medium’s largest entrepreneurship publication followed by +366,835 people.
Subscribe to receive our top stories here.


