profile-img

Artificial Intelligence in the hands of Change Makers

Home  »  Blogs  »  Artificial Intelligence in the hands of Change Makers

Artificial Intelligence in the hands of Change Makers

awards
Sharmistha Dasgupta
Deputy Director General & Head, CoE in AI
sharmi@nic.in

Let us ask what inspires us, our life, our family, hobbies, making money, taking an international flight, travelling to exotic destinations? Well, did anybody think of ‘work’? OK but what type of work, not the mundane go-to office, process the file, do some gup-shup and go back home. That kind of work is OK for a living; but the ‘inspiration’ to work, where do we get it.

What inspires us to work better, work smarter, make our world livelier, make the most out of available resources? Do we let the resources lie around, going about aimlessly through the day, puzzled about what we need to focus on? Well, if you are one of those who consider that they have limited time to make a change in the world and who need to make the maximum out of the resources at their disposal, then you are at the right place. We are here to discuss the treasure trove of information lying around us without being harnessed.

Yes, we can recognize ourselves as the ‘Change-Makers’ and we want to explore Tools and Technologies that will help us move faster and effectively towards our goals. Let us start with one of the emerging technologies – Artificial Intelligence.

Artificial Intelligence (AI) by the term itself suggests that we are trying to emulate human intelligence by getting the work done through some type of system which will try and learn the premise of usecases. It will improve over time to predict the hypothesis so that it is closer and closer to ground truth.

Suppose you teach a child to distinguish between an apple and a guava, he will learn to do so. But if you show him an Orange, he will get confused. Similarly, AI as a technology has learnt to grasp what has been taught, but it is not aware of the context. If you don’t teach the child that orange is also a fruit, he won’t learn.

With this simple clarity on what AI is and what it’s not; we can now delve into how we can harness AI in real world to augment our decision-making capabilities. And identify the work that we can get done using AI.

Typically, in the Government, we are dealing with Mission Mode Projects providing services to citizens all over the country. We have come a long way with digitalization and creating multi tenancy architecture that has created a large pool of data assets; some structured, some unstructured like images of assets created, beneficiaries, responses on the social media etc. Let us consider the popular usecase of VAHAN & SARATHI application of Ministry of Road Transport & Highways (MRoTH). Here lakhs of citizens daily apply to create Learners’ or permanent Driving Licenses. Their first query is how to go about it. Different states, different requirements but a standard interface for applicants, how can we automate the queries of millions of citizens on daily basis. We can have an elaborate help desk managed by human resources or we can have a standard bot that answers most of these queries and a fallback arrangement for a minimal human assisted helpdesk if the bot cannot answer any typical queries. Here comes Conversational AI in picture, bots are answering 2 to 3 lakh queries per day for the RTO, leaving the hands free to deal with other works.

Another smarter way for VAHAN & SARATHI application is to use latest AI technologies for automatic face verification. The photo uploaded by applicant at time of submitting application for Learners’ or Permanent Driving License can be checked while taking the exam through simple face verification. This will reduce chances of fraud by distinguishing clearly if somebody else is taking the exam other than the actual applicant. This is currently being facilitated by AI models through container based micro services architecture and API based ecosystem.

Similarly, Traffic Police enters the vehicle registration number for traffic challan and the driver pays the fine digitally without visiting the court. This can be further fine-tuned if automatic vehicle number plate recognition system is placed at traffic lights. The number plate recognition system can also be used for GSTN to augment intelligence on tax fraud analytics being undertaken on structured data by machine learning techniques.

We can see how multiple AI interventions in our work increases the scope of greater transparency, efficient collections, smarter processes with minimal human intrusions.

National Informatics Centre (NIC) established the Centre of Excellence in Artificial Intelligence (COE-AI@NIC) in January 2019 to help change-makers whether at decision making level or technology implementation level, to come together on the same platform. With a focus on the areas of Machine Learning, Natural Language Processing (NLP), and AI in Text, Image Recognition, Speech Recognition & Synthesis, Deep Learning and Chatbot/ Voicebots, AI Service is provided through Meghraj Cloud of NIC.

NIC also provides AI product services that include:

• AI – VANI: Conversational AI services in form of multilingual chatbots and Bilingual voice support services;
• AI – Satyapikaanan: Face Verification services for contactless mobile based office attendance, Mobile app-based Verification of Life Certificate of Pensioners etc;
• AI – Parichay: ID Verification services for 7 major ID cards for text extraction and matching with online form details;
• AI Manthan: AI Development platform as a Service to interested technical users who would like to build an AI model for some eGovernance projects;
• AI Tainaatee: AI Inference Platform as a Service for deploying AI Models built through AI Manthan
error: Content is protected !!