AAKHYA SINGH

Developed a Deep Learning model that predicts next words while writing articles and course content in real time. The model is built with LSTM (Long short-term memory networks). Created a model that identifies similar asked questions whenever a user asks a new question on the discussion forum. Used the concept of sentence similarity to build the model. Created a Chatbot that addresses user queries, provides course suggestions and handles other tasks while remembering previous conversations. It can also handle off-topic conversations. The bot can currently understand more than 30 intents and can be scaled to include more intents with a small training data of less than 15 examples per intent. The Chatbot is created using the Rasa framework. A Natural Language Understanding (NLU) model is created using a pipeline with different components for tokenization, feature extraction, entity extraction, intent classification that understands intent from user message. Based on the extracted intent from user message, the bot responds to the user.

Supervised product releases by managing team of developers and content creators. Established a workflow for better communication between teams. Came up with product vision and roadmap, aligning the required departments on clear goals to get things over the finish-line. Planned and modified product features based on market analysis and user feedback. Worked closely with the marketing and sales team to analyze performance and plan marketing campaigns and other promotional activities to increase revenue and improve the brand value. Launch of new platforms and product features and effective planning led to increase in the company’s revenue by 66% in an year. Worked with other departments to plan budget, resources, timeline and strategies for different products.

Built applications based on Artificial Intelligence and came up with new business use cases to integrate these applications. Developed a Conversational Chatbot which answers user queries by automating communication. It understands the context, remembers previous conversations and accordingly addresses user queries. Used Rasa NLU framework to create the Chatbot. The pipeline used has different components for tokenization, feature extraction, entity extraction, intent classification, which can be customized based on the training data. Integrated Sentiment Analysis with the Chatbot to understand the sentiment from user message along with the context and give the response accordingly. Responsible for troubleshooting and handling the issues for the three clients using the Chatbot application. Created a similar records matching service which allows the user to search some text for different entities and displays similar records from the data provided by the user. The approach used here is sentence similarity. Developed a text summarization service which generates summary for the text provided by the user using PageRank algorithm. Designed a model that identifies digits, characters and punctuation marks from handwritten data. Used a Sequential Convolutional Neural Network model built using Keras achieving 90.74 percent accuracy.

CBSE board

CBSE board