top of page
Writer's pictureAbhay Kulkarni

And then there is “Practical AI”

Updated: Apr 4, 2020

Look around. Every one talks about AI. AI as a killer, AI as a savior, AI as an enabler. Whatever the expectations from AI may be, how many AI experiments actually make it into production environments? How many of these AI projects are really successful? The answer might make you want to quit a not-yet-started career in AI. Reality is AI implementations can succeed. But a lot of us work on AI based features as projects, research or as something to please management. There is AI done to create models and test them. And there is AI done, guided by business realities and practical considerations. I term this AI as “Practical AI” - AI done with a rational mindset grounded in reality of IT, business, entities, budget, processes and people.


I have built AI features for a few product lines and led data science and product teams in delivering such features. The key to success is ”keep it real as in keep it practical”. This series of blogs will discuss key aspects that keep ”building AI” practical.


If you do create a log in (and I encourage you to do so), you will get access to two slide decks that explain my approach towards identifying features that could benefit from AI, and steps for developing and implementing AI features.


Also, this is my attempt to share my learning. And I do learn every day. So I do reserve the right to change my mind as the thought process develops. One thing I can promise though: You will only stand to benefit from what you read here. Feel free to agree, disagree and comment. It will only help improve our collective learning.


See you then at the next post! Let’s create Practical AI #abhayPracticalAI #artificialintelligence #ai #machinelearning

11 views0 comments

Recent Posts

See All

Prediction end point - streaming or batch?

We have discussed a few aspects of defining a prediction end point. Here is one more. Always work with your business users to understand...

Defining prediction API - some tips

The end result of a requirement gathering exercise for AI is often the end point definition for the desired prediction rule. Defining the...

Comments


bottom of page