AI without data

AI without data

Artificial intelligence or AI refers to the simulation of human intelligence in machines. (Refer the article Are AI and machine learning the same if you wish to understand more on what is AI) Artificial intelligence-based system learns from data (also referred as “training data”). These systems find relationships hidden in the training data, develop an understanding of the data, make decisions and then evaluate their performance from the data. And better the training data, better the system performs. Or in other words, the training data (both the quality and quantity) dictates how well the AI system will perform. In this article I will cover some hacks one can use if there is not enough training data available. So read on.

But first let’s understand how the training process works in the AI world.

How does the training process work?

Let’s say you want to make an AI-based system capable to do visual search i.e. it should be able to recognize a particular object in a given image (example: is the person in the photo wearing a jacket or not?). The process of training this AI is as follows (refer the picture below)

AI training process

  1. A computer program is fed through a large number, usually thousands, of labelled images of “person wearing jacket” (positive label) and “not wearing jackets” (negative label). This is the training data for the AI.
  2. The program learns on this data and produces a model to detect jackets.
  3. This learning is repeated in fixed iterative cycles till the performance of the model improves and then the model is able to detect the jackets with good precision in new unseen images. This iterative process is called Training.

Businesses often struggle to find enough relevant training data to train their AI and this is because most often the data they need resides in multiple sources, are usually available in an unstructured format and therefore not useful for machines to learn from. So in such situations, how can you still make sure that AI works for you?

What can you do when you don’t have much data?

Here are 5 possible hacks you can try out and it is very likely that one or more will work in your situation.

Top 1 hack: Transfer learning

Transfer learning is about using the knowledge gained while solving one problem for which a lot of labeled data is available and applying it to a different but related problem for which only little labeled data is available.

For example, the knowledge gained while learning to recognize a person in an image could apply to recognize the types of clothes that the person is wearing and finding dominant colors in his clothing.

Top 2 hack: Data augmentation

Data augmentation is a process that increases the diversity of data available for training models, without actually collecting new or more data. As an example, for computer vision-related tasks we have the following options:

  • Horizontal/Vertical Flip: Flip the image horizontally or vertically to add more diverse data.
  • Change in scale: Crop and/or zoom images in various scales
  • Lighting conditions: Change the brightness/sharpness/contrast of the images available.
  • Rotation: Rotate the image by some different degrees.

For text (NLP) related task you can try the following options:

  • Synonym Replacement: Randomly choose some words from the sentence and replace each of these words with one of its synonyms chosen at random. This creates more utterances of as existing sentence.
  • Back Translation: English is one of the languages which is having lots of training data for most of the NLP(Natural Language Processing) tasks. So, if there is less data in a language for a particular task then more training data can be created by translating English data into that language.

Top 3 hack: Problem Reduction

Problem reduction approach refers to modifying the new data or unknown problem to a known problem such that it can be easily solved using existing techniques. Here is an example:

  • Suppose we have a small labeled dataset of “audio clips”, where the sound in each clip belongs to one of the 10 categories like a car engine, piano note, drilling machine, and a dog barking. Now say, we want an AI-based system capable of recognizing the source of audio in these clips.
  • If you try using any existing audio-based algorithms to solve this problem, it might work out. However, the small data set can be a deal-breaker.
  • One alternate method is to try converting the “audio clips” intoimages” thereby reducing the audio problem to an image classification problem. Now we can use any suitable computer vision-based algorithm to train on this “new image” dataset and create a system capable of classifying these “audio” images into one of the 10 categories.
  • Surprisingly results have shown that the performance of these image-based systems is far better than the audio-based systems.

Top 4 hack: One-shot learning

Neural networks are good at learning from high dimensional data like images or spoken language, but only when they have huge amounts of labeled examples to train on. Humans, on the other hand, are capable of one-shot learning – if you take a human who has never seen a fork in his entire life and show him one, next time if you ask him to bring one from that messy drawer of your kitchen, he might be able to fetch that with high precision. For this type of approach, there are advanced algorithms that we will try to cover in later blogs.

Top 5 hack: Open source datasets and APIs

Various data sets are available in the open-source community or on the internet, which can be easily downloaded and used for training purposes. Some examples like the amazon dataset and google dataset are also accessible easily. Apart from these, search engines like google and bing also provide APIs to download datasets and pre-trained models, which can be used in many ways to improve the performance of an AI-based system.


While availability of relevant training data decides whether an AI system will perform well or not, it is certainly not a hopeless case if one does not have enough data. There are multiple techniques and hacks that one can still try out and I have listed the top5 above.

Where chatbots can be used? Who uses chatbots?

Where chatbots can be used? Who uses chatbots?

Chatbot technology have been around for years now. They started out as an accident on Telegram and then developers started creating little bots on Slack – internally for their team’s use. Then WeChat, China’s Biggest Messaging Platform, began supporting bots and it really took off in China. Since Facebook Messenger, WhatsApp, Skype, Slack, and a growing number of bot-creation platforms came online, developers have been churning out chatbots across industries with Facebook’s most recent bot count standing at over 33,000.

With chatbots then becoming a serious business, the question comes – where can chatbots be used? Who uses chatbots?

Here are the top 7 chatbot usecases and application areas for businesses.

Consumer Engagement at Scale:

Brands that engage with their customers on social media platforms make 20% to 40% more money, according to Bain & Company. Nowadays, most businesses have an online and social presence. But since availability of manpower 24/7 is a constraint and comes at a cost, it is not always possible to keep the engagement channel open and active at all times. Businesses can use chatbots as their proxy agents to engage customers at scale 24/7 and across channels such as web, mobile, facebook, whatsapp etc. Chatbots, usually, only give out a slice of information at a time and can guide the interaction based on the inputs provided by the user. That way the engagement becomes a lot more interactive and chances of losing a customer is also a lot lower.

Improve Sales:

Chatbots make the lead generation process faster & easier by interacting & engaging customers in every step. From the moment a customer makes first contact on a business’s website or social handle, a chatbot can engage the customer by understanding her requirement, recommending a product/service and guiding her to the next step in the buying process. With AI powered self-learning models in the background, chatbots can even be trained to make complex decisions such as sending coupons to customers so that they will come to the store and purchase.

Marketing campaigns:

Chatbot technology can be very useful to marketers within an organization – instead of spending money on banner ads which are typically one-way communication, marketers can leverage the chatbot technology and convert their banner ads into a two-way, highly interactive and personalized communication channel with their customers. In the marketing world it is referred to as ‘conversational ads’ and typically these result in improved click-through-rates and higher ROI compared to traditional advertisement methods.


Finance function within a company is often very complex, involves lot of stakeholders and prone to repetitive tasks such as generating same report multiple times or filling up the same forms again and again. A conversational bot can be very useful in this situation leading to not just process improvement but also saving significant money. Here are a couple of use cases that can deliver instant ROI

  • Accounts Payable Virtual Assistant – Your suppliers have submitted invoices and are now chasing you for the payment. Often they will call your team, or send email multiple times and every time your team will check the status and will write back to them with the latest update. An AI bot can take over the entire process from your team and allow your team to focus on the more strategic pieces. Let your suppliers ask their questions to the ai bot and let the chatbot answer them back instantaneously.
  • Expense Filing and approval – Every employee in every company abhors expense claim process since it is mundane and extremely time consuming. It involves filling up the same form every time and then wait for the internal approval workflow to settle the expense. An AI bot can not only automate this entire process, but can also make the experience of expense filing a lot more enjoyable for your employees


The chatbot technology can benefit e-commerce shop owners as it allows them to have a Virtual assistant available always on the Webshop for helping visiting clients find the right product/service quickly. Reports suggest that web shops with virtual assistance service can not only convert 10X more than normal websites, but it also leaves a much more delightful experience for the visiting customers prompting them to come back again for the same personalized service and experience.

Chatbots Customer Service:

70% of questions that customer service agents answer over calls are repetitive in nature i.e. these questions are largely order related or delivery related and therefore are great candidates for automation through an intelligent chatbot. Your customers get instant answers without having to wait in the queue to talk to your agent while you get to save a lot of money by making sure your agents are now able to handle 2X-3X more calls within the same time.

Collect Customer Data & Gaining Insights:

As stated earlier, chatbots are great tools to interact with customers. From the feedback that chatbots collect by asking simple questions, when analyzed carefully, a brand can make improvements in their service or product. Moreover, the insights gained by chatbots can also help a brand optimize their low converting webpages. For instance, if a landing page brings in massive organic traffic but doesn’t convert well, chatbots can reach out to customers to collect information about why they are leaving the page without making a purchase.


Chatbot technology is opening up a world of new opportunities for businesses across industries and sectors. There will be innumerable chatbot usecases where these intelligent assistants can be deployed and in the above paragraphs I have just highlighted a few of those use cases to get you started. The only advice I will leave you with is that always start with the business problem you want to solve and then evaluate whether a chatbot is the right choice to solve that problem. If yes, proceed or else stop and look for another problem.

Curious much? – Mail me at and let’s start a conversation.