Now Reading
Where’s The Money In Artificial Intelligence? Find out areas that have proven the money value in AI

Where’s The Money In Artificial Intelligence? Find out areas that have proven the money value in AI

Richa Bhatia

Despite the stupendous leap in advances in artificial intelligence and machine learning, the underlying sentiment has always been about getting a formidable return on investment. None of these advances would matter if it doesn’t give a robust ROI. And all the cheery news about AI startup acquisitions point to one thing – tech behemoths are in the game for reaping money in the long run. At present, Google has made the maximum investment in AI acquisitions. So how does one make AI useful? Well, if you are a startup you need to do your market due diligence and find out what the customer wants and the needs that can be addressed through AI.

It is still early days in AI and most enterprises and startups are struggling to find a good business model for AI or machine learning.

In this article, Analytics India Magazine shines a spotlight on ways AI startups across the globe and in India are making money by presenting real-world examples of how startups are extending the capabilities of AI:

1. Making AI profitable with Structured Data: Palo Alto startup Diffbot founded by Mike Tung becomes the Google of structured data. Defined as the key to AI, structured data has proven its money value with startups and financial enterprises investing heavily in structured data. According to Palo Alto-based structured data startup founder Mike Tung, cash-rich digital natives such as Google, Baidu and Amazon have the human muscle to spider web and gather data, small companies lack the tech muscle and an inventory of structured data to build a database of structured knowledge. So great has been the response to Diffbot, founded in 2010 that the company’s taxonomy contains 1.2 billion objects and adds 10 million objects per day. As a matter of fact, Tung reportedly revealed that Google’s Knowledge Graph recently crossed one billion objects. Tung, a Stanford dropout is tackling the biggest problem startups and other companies face today – presenting data in a structured way so that AI systems can read it. The startup has some bold-faced names as clients listed in its roster – Yandex, eBay, Cisco, Adobe and even AOL.

Upside: With structured data being the key to achieving business objective, this could be a way for companies/startups to make data useful for small and medium businesses

Downside: No proven business model in India

Who your customers would be: Mid-size to large consumer product companies who want to tap into what their customers want

2.Early Bot builders are minting money: Bot business is booming in India. Think and Singapore-based which are the talk of town, flushed with VC money and a great customer base. According to chatbot developer Ekim Kaya, chatbots are a popular go-to strategy for AI startups since they have the potential to scale and have a clear business model – there is the SaaS Bot model for B2B and B2C SMBs and big enterprises alike. Kaya believes as the market matures, the profitability will increase and founders will find new ways to monetize bots besides developing APIs. He emphasizes the rise of “Bot Tester” and “Bot Conversational Flow Designer” in the future.


  • You don’t have to develop a separate app, one can run it in a messenger Facebook or Slack
  • Another major advantage is that this has a proven business model and VCs have shown a great interest in chatbots
  • The go-to business model is that of Software as a Service (SaaS), and is delivered as an API
  • While there is the traditional business model of ad-selling on bots peddling native and sponsored content, data gathering can also become another revenue generator for chatbots that are a great tool for collecting user data (location, email, preferences) and interests. This data can help brands understand the audience better

Downside: The market is flooded with chatbots that are being deployed by big enterprises to minimize front line customer service support staff and handle low-level queries. Right now, there is way too much competition in the bot world. To build a profitable chatbot one must understand the user psyche well to ensure that the end product is not deemed spammy.

Who your customers would be: Well, the most successful use cases are coming out of the finance, e-commerce, healthcare and retail sector wherein bots are automating simple tasks and have increased operational efficiency.

3. Customized solutions like image recognition & text recognition: Over the years, compelling AI-optimized use cases have emerged for areas such as image recognition, text analysis, speech to text conversion that can deliver better outcomes. While tech giants Google, Microsoft and Apple are integrating AI across their user interfaces, such as personalizing search and marketing, small businesses that lack the economic muscle are benefitting from AI-powered forecasting that is driving marketing operations and personalization. Case in point – that rolled out its AI-powered product AIDA, aimed at marketers to help personalize their customer touch, thereby improving engagement, increasing transactions and reducing attrition. Another Bangalore-based startup Artifacia has an AI-powered visual discovery platform that is revolutionizing visual search for e-commerce companies.

See Also
How SMBs Can Maximise Output From Their Small Data Science Team

Upside: With more and more value-adding use cases emerging over the years and increased adoption of AI at scale, budding startup enthusiasts can identify the problem areas they want to focus on and build capabilities around it.

Downside: There is always the fear of crashing out and finding not enough VC interest to sustain.

Who your customers would be: Right now, the sectors that are showing the most investor are financial services, high tech and telecommunications, automotive, followed by healthcare, retail, CPG and education.

4. Selling Cognitive Software:  Here’s a thought, if Watson open sources its APIs in the near future, this could lead to customized AI solutions. The commoditization of cognitive tools could lead to more AI-focused solutions across services (forecasting, optimization) and products (semiconductors). If the developers starting using a vendor’s open-source cognitive tools, this could also help in introducing a standard AI software. Right now, Google’s Cloud ML has taken concrete step to allow programmers create solutions via Tensorflow.

Upside: In India, use cases are still weak.

Downside: One will have to follow a Follow a test and learn approach

Who your customers would be: Given the significant advances in artificial intelligence, cognitive solutions can benefit all sectors, particularly healthcare, retail, education and even manufacturing.

What Do You Think?

If you loved this story, do join our Telegram Community.

Also, you can write for us and be one of the 500+ experts who have contributed stories at AIM. Share your nominations here.

Copyright Analytics India Magazine Pvt Ltd

Scroll To Top