Projects

Projects

🚀 My Projects

🗣️ Real-Time AI Voice Agent for Telecom Outbound and Inbound Calls

📅 2025 | 🏢 Innovation Incubator Advisory

  • Focusing on a Real-Time Voice Agent system for outbound and inbound telecom communication leveraging Livekit to initiate and manage real-time audio calls to users.
  • Integrated Amazon Nova Sonic speech to speech model to dynamically generate conversational responses during calls. Used knowledge base for real time retrieval of telecom data.
  • System can deliver information and interacts live with users to resolve follow-up queries. Designed voice workflows with prompt engineering for natural, human-like interaction tailored to telecom use cases.
  • Technologies: Python, Livekit, Amazon Bedrock, Amazon Chime, Twilio, Boto3, FastAPI.
  • 🔗 Article

🤖 AI-Powered Telecom Chatbot

📅 2024-25 | 🏢 Innovation Incubator Advisory

  • Focusing on developing an AI-powered telecom chatbot to automate customer support and query resolution.
  • Implemented intent classification and entity recognition using Amazon Lex. Integrated knowledge-based real-time telecom data retrieval using AWS Bedrock, and optimized response generation with prompt engineering.
  • Technologies: Python, Amazon Lex, Amazon Bedrock, Boto3, Django.

🧠 Multi-Agent AI System for Telecom Automation

📅 2024-25 | 🏢 Innovation Incubator Advisory

  • Focusing on the development of a multi-agent AI system for telecom operations.
  • Designing intelligent agents to handle diverse tasks, including customer queries, API interactions, and knowledge retrieval. Optimized agent workflows and improved system efficiency for automated telecom support.
  • Technologies: Python, Amazon Bedrock, Boto3, Django.

🗣️ AI Voice Agent for Outbound and Inbound Calls with AWS Bedrock, Lambda, Lex and Connect

📅 2024 | 🏢 Innovation Incubator Advisory

  • Focused on a Voice Agent system for outbound and inbound telecom communication leveraging AWS Connect, Lex and Lambda to initiate and manage real-time audio calls to users.
  • Integrated Amazon Bedrock Agents with anthropic models to dynamically generate conversational responses during calls. Used knowledge base for real time retrieval of telecom data.
  • System can deliver information and interacts live with users to resolve follow-up queries. Used prompt engineering in Bedrock agents for natural, human-like interaction tailored to telecom use cases.
  • Technologies: Python, AWS Bedrock, AWS Lex, AWS Connect, AWS Lambda, Boto3.
  • 🔗 Article

🗂️ Appraisal Digitization

📅 2024 | 🏢 Innovation Incubator Advisory

  • Focusing on automating the digitization of appraisal documents.
  • Developed a classification model and conducted data extraction using OCR techniques, leveraging prompt engineering for efficient extraction.
  • Technologies: Python, Scikit-learn, SpaCy, NLTK, Tesseract OCR, Paddle OCR, Google Cloud Vision OCR, OpenCV, OpenAI, Gemini, AWS Textract, FastAPI.
  • 🔗 Project Files

🌾 Citation Network Analysis of Robotics in Agriculture

📅 2023 | 🏢 Digital University Kerala

  • Conducted Citation Network Analysis on the topic “Robotics in Agriculture” using data from Web of Science database.
  • Proficient in Pajek and Gephi for main path and cluster analysis. Utilized NLP techniques for word frequency analysis with bar plots and thematic clustering using word clouds.
  • Identified developmental trajectories of agricultural technologies and the growth of technological advancements in the sector.
  • Technologies: Python, Gephi, Pajek, Matplotlib, Seaborn, Word cloud, NLTK.
  • 🔗 Project Files

📸 Real-time Attendance Marking System with Automatic Face Recognition using TensorFlow Lite

📅 2023 | 🏢 Digital University Kerala

  • Created a facial image dataset of students. Used Label Studio for annotation and Roboflow for dataset enhancement.
  • Implemented object detection using a pretrained SSD MobileNetV2 FPN-Lite 320 model to develop a TensorFlow Lite model for face recognition.
  • Integrated the model for real-time attendance tracking, updating an Excel sheet based on detected faces via Computer Vision, reducing manual effort.
  • Technologies: Python, Label Studio, Roboflow, TensorFlow Lite.
  • 🔗 Project Files

🎬 Multitasking App for Tumor Detection, SMS Spam Detection and Movie Sentiment Analysis with Deep Learning

📅 2023 | 🏢 Digital University Kerala

  • Developed a Streamlit app for:
    • 🧠 Tumor Detection using CNN
    • 📩 SMS Spam Detection using RNN, LSTM
    • 🎞️ Movie Sentiment Analysis using GRU
  • Utilized TensorFlow for model implementation. Created an interactive UI using Streamlit for enhanced user experience.
  • Technologies: Python, TensorFlow, Streamlit.
  • 🔗 Project Files
  • 🧪 Demo

🗑️ Smart Dustbin using Arduino

📅 2022 | 🏢 Digital University Kerala

  • Designed and executed a Smart Dustbin prototype featuring Arduino, UV sensor, and Servo motor integration.
    Connection Smart Dustbin project
  • Implemented ultrasonic sensing for automated lid opening/closing based on proximity. Aimed to enhance hygienic waste disposal by reducing manual contact.
  • Technologies: Arduino Uno.
  • 🔗 Project Files