Innovating @ Scale with AI

AI is here and now. Enterprises that want to succeed and be counted as leaders cannot ignore AI any longer. I guess this much is clear.

But I repeatedly see organizations struggle with using the power of AI to innovate fast and at scale. The first thing a lot of them do is hire a bunch of "data scientists" and they all very quickly start working on building the greatest algorithm - but very quickly it all starts going downhill. The performance of the algorithm never seems to quite live up to the initial hype. The business leadership losses interest and within a year there is disillusionment everywhere.

Nobody is quite able to see the light at the end of this tunnel.

But it does not have to be like this. Most of building an AI solution is grunt work that can and should be automated. The first step should be automate the AI pipeline as much as possible. There are enough tools in the market (both open source and commercial) that can help do this. Once this is in place, the AI algorithm even if it is not completely ready should be exposed to the business and their feedback solicited. Early feedback from "business" is pure gold and it helps establish the much needed "trust" between the business and data science teams. Now the algorithm can be tuned and trained with larger sets of data in parallel to provide improved version(s) of the product at regular intervals.

This approach lets the business team(s) drive the development of the AI infused product and gives them a sense of ownership which is critical to the success of the product / project.

Once again, for the data science team, automate as much of the pipeline as early as possible and expose the unfinished algorithm to the business users sooner than later. Then iterate the product / project in partnership with the business. 

This is the approach that will separate teams that use AI to accelerate their companies growth from those who will fall behind. 

To understand more and to see how we can help your organization leapfrog to the next level using AI, please visit us @ www.roundsqr.com.

Dr. Chandra Bhushan K, PhD

Management Consultant , Advisor - Strategy, Technology, Operations

6 年

AI may also go through the same life cycle that of web development in terms of reaching a maturity level in the enterprises. Looping feedback automatically into the system can speed up. Enterprises? may not like to get in R&D mode for AI implementations as it is too expensive and time consuming. The enterprise product companies may need to adopt? integrated AI features for the enterprises to put into practice.?

要查看或添加评论,请登录

Srinivas Atreya的更多文章

  • Well-being in an age of abundance

    Well-being in an age of abundance

    Our journey of about 200,000 years on this planet can simply be understood in terms of tools and techniques employed by…

    3 条评论
  • The Data Scientist is dead! Long live the Data Scientist

    The Data Scientist is dead! Long live the Data Scientist

    One of the questions that I get repeatedly nowadays from a lot of beginners in the data science space is around the…

    1 条评论
  • The softer costs of being a Data Scientist

    The softer costs of being a Data Scientist

    Data Science is hot. Everyone wants to learn and become a Data Scientist.

    1 条评论
  • Democratizing ML in the Enterprise

    Democratizing ML in the Enterprise

    ML is everywhere now and it is an accepted truism that enterprises that harness it better will lead the future. Now…

    1 条评论
  • The Other Side

    The Other Side

    Now that the world has effectively shut down with no clear indication of when things will return to normal, I've…

  • Making Neural Networks Useful

    Making Neural Networks Useful

    It seems that neural networks are eating the business world for breakfast nowadays. I can hardly get through a day…

    2 条评论
  • Looking Back, Looking Ahead

    Looking Back, Looking Ahead

    As I look back on 2019, I am genuinely surprised. I, with some of my friends started a Data and ML company (www.

    4 条评论
  • To New Beginnings

    To New Beginnings

    It is that time of the year again. To introspect on the year that went by and make plans for the new one just around…

    2 条评论
  • Moving an ML model to production is not the time to disband your team!!

    Moving an ML model to production is not the time to disband your team!!

    One magical thing about software is that once it’s deployed, it keeps working. Once an accounting software is put in…

    2 条评论
  • Auto Drivers and Corporate Strategy

    Auto Drivers and Corporate Strategy

    I live and work in Hyderabad, India and commute everyday to an area called Hi-Tech City which is the hub of a lot Tech…

    1 条评论

社区洞察

其他会员也浏览了