For meaningful and actionable insights, data has to be structured
Published by : Industrial Automation
Neelakanth Veluru – Product Manager, teX.ai.
What was the rationale for launching teX.ai, and why a SaaS based model?
Irrespective of your company's performance, it is imperative to analyse all available data to stay competitive in the market. Tremendous volume and variety of data is generated every day in the form of text data. This data is highly unstructured available and is available in the form of images, scanned documents, digital documents, voice and video formats. It however contains important information that could help ameliorate businesses strategies and fulfil customer expectations (both internal and external). We have to structure the data to derive meaningful and actionable insights. Using manual effort to derive these insights proves to be – time consuming, requires a greater number of resources and proves to be very expensive. teX.ai was created to solve these business problems.
SaaS model has become a preferred way of consuming different services for many organisations. It enables organisations to shift expenses from CAPEX to OPEX and also provides flexibility and scalability. SaaS model also helps organisations to focus more on their core activities by relieving them from managing and maintaining software.
How different is AI-based text analytics from data analytics? What are the target areas?
Bring on the text is our motto and text data is our focus. Text data is completely unstructured and very complex to analyse whereas quantitative data is easy to work with and is kind of a matured field. Our key focus areas are across industries such as retail, insurance, manufacturing, banking and more.
How does the solution work in practice – the logic and the analytics?
There are three phases in text analytics:
1. First, we ingest data from disparate sources such as documents, images, social media and many more
2. Secondly, we have to extract and convert this data into a structured format, and
3. We then apply analytics on this structured data to derive insights.
This is the generic high-level approach that we follow, and it is completely automated with the help of teX.ai.
While one can see the potential in service industries, how useful will this be in the manufacturing or process industries that are already facing data deluge?
Major problems that manufacturing industries face with data are:
1. Volume of data – Handling it manually is very difficult and complex
2. Existing systems cannot analyse unstructured text data, so the need for newer technologies arises, and
3. Not all unstructured data is relevant.
Hence, to tackle these problems, we have to use software products powered by Artificial Intelligence technologies such as Machine Learning (ML), Natural Language Processing (NLP) and more.
Data analytics today is hot as career option, and presumably facing skills shortage. What is the situation on ground?
A career in data analytics or data science is like two sides of a coin. One side being where the candidate should possess functional skills such as business acumen, domain expertise and should be able to come up with innovative business use cases. The other side of the coin being technical expertise where the candidate should possess tool and technology expertise, hands on experience, willingness to be flexible and should possess a high learnability index. Additionally, Data Scientists need strong skills in Math and Sciences to be able to implement various algorithms and packages needed.
How big the potential market for text analytics services is in India and globally?
The global text analytics market is expected to reach US$ 15 billion by 2025. Even in India we see a significant rise in the need for text analytics. Every need is not the same. There are a few common use cases across industries, but text extraction and analysis are unique to each enterprise's requirement. With businesses adopting newer technologies, their demands are becoming more and more advanced. A few use cases have never even been seen before. This tells us that the growth of the text analytics market may even double after 2025.