SRAVAN REDDY
Applied ML and Computer Vision Engineer
Specialized in Computer Vision, NLP and Deep Learning
Explore My WorkAbout Me
B.Tech student in Computers and Communication Engineering at Amrita Vishwavidyapeetham (2021-2025). Passionate about AI, machine learning, and IoT, with innovative projects including surgical assistance systems and crop prediction models.
Skilled in Python, computer vision, and NLP with hands-on experience in developing speech-to-text systems, RAG models, and voice-controlled applications. Currently interning at Loksun.ai to enhance expertise in ML and AI. Holds certifications in Google Data Analytics and Python for Data Science, reflecting commitment to data-driven solutions. Aims to leverage technology to solve real-world challenges in AI and IoT.
Experience
- Developed a multi-Indian language speech-to-text system using Vosk for offline processing and Whisper for GPU-powered transcription, automating manual transcription tasks
- Built a RAG model for the Indian Constitution using LangChain and Gemini 1.5 Flash, delivering accurate, context-aware legal responses
Education
Specialized in Computer Science with focus on AI, Machine Learning, and IoT. Completed coursework in Data Structures, Algorithms, Computer Vision, Natural Language Processing, and Deep Learning. Actively participated in research projects and technical competitions.
Projects
Voice-Controlled Surgical Assistance System
Engineered a voice-controlled system integrating Google Speech-to-Text and YOLOv11 for surgical instrument detection with 97% accuracy. Developed a comprehensive dataset of surgical tools containing over 6,000 original images, training specialized models that reduced command misinterpretation by 40% in clinical environments.
Jan 2024 - May 2024
Crop Prediction
Built predictive models for 22 Indian crops using historical data, achieving 20% reduction in prediction error. Enhanced model accuracy by implementing advanced data preprocessing techniques, including feature scaling, outlier detection, and dimensionality reduction.
Aug 2023 - Nov 2023
Health Monitoring System
Designed an IoT-based system to monitor temperature, pulse, and BMI using Arduino and ThingSpeak, improving data accessibility by 75%. Integrated ThingSpeak cloud IoT platform to enable live uploading and remote access of patient health data, ensuring authorized personnel can monitor real-time health conditions from any location.
Feb 2022 - May 2022
Speech-to-Text System (Indian Languages)
Developed a multilingual speech-to-text system using Vosk for offline transcription and Whisper for GPU-accelerated processing. Automated transcription workflows across Indian languages, improving speed and accuracy while supporting dialect variations.
May 2024 - Jul 2024
View on GitHubRAG on Indian Constitution
Built a Retrieval-Augmented Generation (RAG) model using LangChain and Gemini 1.5 Flash to answer constitutional queries. Combined contextual retrieval with generative reasoning for accurate legal insights.
Feb 2024 - Apr 2024
View on GitHubSurgical Tools Dataset
Created a comprehensive dataset of 6,000 high-quality images for surgical tool recognition, including 9 categories of tools. Captured under diverse conditions with overlapping configurations to train robust computer vision models for clinical applications.
Jan 2025 - Feb 2025
View Dataset
LinkedIn