ANKIT SINGH

Accomplished AI Developer with around 7 years of experience, specializing in AI, Natural Language Processing, and machine learning. My expertise includes developing innovative solutions and deploying data-driven applications across various platforms.

• Introduction to TensorFlow for Artificial Intelligence, Machine Learning, and Deep Learning
• Neural Networks and Deep Learning

• 91-7000660377, India

KAIT TECHNOLOGIES (KUWAIT)

CHAT WITH DOCUMENTS(CHATGPT)

• Used Open AI and LangChain for creating custom chatbot on client data.
• Created vector database for context and generating responses using OpenAI Embeddings and LLM's.
• Created Input and Output Parsers and Evaluation data using LangChain api's.
• Integrated with Whatsapp Api for chat on Whatsapp platform.

HuggingFace Projects.

• Used Hugging Face open source models for generating text to images.
• Created a model for text summarization using hugging face.
• Used hugging face model for Name Entity Recognition(NER) in text.
• Created a Chatbot on custom data using hugging face Falcon-7B model.

AI Chatbots

• Developed Chatbots on Multiple Platforms (Whatsapp, Instagram, Google messages, Twitter, Fb messenger) for various domains and languages using RASA API's.
• Implemented intent classification and entity extraction using RASA.
• Developed and deployed on AWS servers.
• Created chatbots in Amazon Lex, Google DialogFlow, and Microsoft bot.

LOGICSOFT - BEYOND HYPE (GURUGRAM)

Importer/Exporter TaxId Prediction

• Predicted Importer/Exporter Tax-ids using deep learning models on Tensorflow and Keras.
• Achieved 95%+ accuracy and deployed models on Azure server.
• Configured GPU server machines for Tensorflow-gpu.

HsCode Prediction (Text Classification)

• Directorate General of Commercial Intelligence and Statistics (DGCIS),
• Developed 99 deep learning models chapter-wise on Trade data for predicting Hs-codes.
• Integrated with MySQL Database for data upload/download.
• Achieved approximately 90% accuracy in each model.

BIGAPP TECHNOLOGIES (BANGLORE)

AI ChatBot

• Developed an AI-driven Chatbot in Python for domain queries of Flights, trains, and cabs.
• Implemented word embeddings using Word2vec and Glove.
• Built generative model using RNN (BI-LSTM) for seq2seq generation.
• Implemented Tf - Idf technique for retrieval-based model.

Sentiment Classifier

• Built a Sentiment Classifier integrated into the chatbot using NBSVM.
• Collected Data from Twitter Using Hadoop and Flume.
• Achieved 90% accuracy in text classification.

PRAPYA (BANGLORE)

Student Grade Prediction

• Developed a classification model to be deployed in schools for student grade classification.
• Used Logistic regression and achieved 87% accuracy.

Loan Repayment

• Developed a Machine learning model for predicting Loan repayment chances of students.
• Analyzed important factors for prediction.

Patient Readmission

• Developed a prediction model for a small hospital to predict patient readmission probabilities.