Riya Kerur
Embedded Systems | IoT Solutions | Real-Time Systems | Computer Vision
📄 Download Resume
Embedded Systems | IoT Solutions | Real-Time Systems | Computer Vision
📄 Download ResumeImplemented WebP image format across front-end assets, reducing bandwidth usage and improving mobile performance. Built a modular customer analytics dashboard using Flask (Python) for logic and Spring Boot (Java) for backend services, with integrated RESTful APIs and real-time database operations. Defined and implemented technical specifications: designed database schemas, set access control rules, and documented RESTful interfaces, resulting in faster deployment cycles.
Identified high-risk credit customers using unsupervised K-means clustering on 150k+ customer records, driving more informed risk segmentation. Delivered actionable insights to stakeholders by analyzing data trends with Python (Pandas, NumPy) and presenting clear EDA visuals. Recommended targeted marketing strategies based on model-driven insights with 87% clustering accuracy.
Installed network switches and configured IP-based CCTV surveillance systems across Tata Motors’ assembly line after conducting on-site evaluations. Integrated IP cameras with Network Video Recorders (NVRs) via VLAN and SNMP routing, enabling secure, real-time mobile access. Strengthened remote surveillance security by implementing IP whitelisting and multifactor authentication, boosting system integrity by 15%.
I’m a graduate student pursuing my Masters in Computer Engineering at California State University, Sacramento. I specialize in building intelligent, real-time systems at the intersection of software, hardware, and data. My experience spans software development, embedded systems and computer vision. As a Junior Developer at Arcasan Software, I optimized front-end performance and built scalable analytics dashboards using Flask, Java, and REST APIs. At SmartKnower, I developed machine learning models to uncover trends in large-scale financial data and supported strategic decisions through data analysis. During my time at Visual Securities, I configured secure IP-based surveillance networks for industrial settings, gaining hands-on exposure to networking and IoT infrastructure. I’m skilled in tools and languages like Python, C++, Java, Flask, SQL, and frameworks including Streamlit, Dash, Scikit-learn, and Dlib. My interests lie in computer vision, edge computing, and intelligent IoT systems, and I’m always looking for opportunities where I can blend creativity with engineering.
Pandas, NumPy, Scikit-learn, Dlib, Streamlit, Whisper
Arduino, Raspberry Pi, MQTT Protocol, Real-Time Alert Systems
OpenCV, Tensorflow, PyTorch, Facial Recognition, Image Classification
April 2025 – May 2025
Built a Streamlit-based web app that summarizes YouTube videos using Google Gemini LLM with Whisper as a fallback for audio transcription. Integrated a dual-mode transcription system to ensure complete video coverage, switching between YouTube captions and Whisper-based local transcription. Implemented input validation, text preprocessing, and prompt chunking to handle longer transcripts and ensure accurate summarization. Achieved real-time, multi-paragraph summaries through seamless Gemini API integration and robust NLP pipeline. Demonstrated full-stack development skills, combining LLMs, audio processing, and UI/UX design for end-to-end functionality.
Jan 2025 – Feb 2025
Designed and implemented a smart 4-way traffic controller that dynamically adjusts signal timings based on real-time vehicle density. Programmed adaptive traffic logic in Embedded C using Arduino IDE; integrated HC-SR04 ultrasonic sensors to detect vehicle presence. Developed signal-skipping algorithms to minimize idle time and optimize vehicle flow through intersections. Enabled real-time remote monitoring and control by integrating the system with the Blynk IoT platform. Improved traffic efficiency and responsiveness while demonstrating end-to-end IoT solution development skills.
July 2022 – Feb 2023
Designed a dual-authentication system by integrating RFID-based user verification with facial recognition, aimed at reducing unauthorized access in physical security environments. Programmed facial recognition using Python, Dlib, and CNN embeddings, enabling the system to detect and encode facial features with around 78% accuracy, even under variable lighting and angles. Deployed the solution on Raspberry Pi, ensuring fast and reliable edge processing without needing cloud support, and successfully demonstrated improved security performance compared to single-factor authentication systems.
Jan 2022 – May 2022
Designed and implemented a secure communication protocol enabling real-time data exchange between distributed sensors and a central monitoring system, incorporating end-to-end encryption to ensure data confidentiality and integrity. Developed and trained an Isolation Forest anomaly detection model using Python and scikit-learn, successfully reducing false positive rates by 30%. Created a user-friendly monitoring dashboard with OAuth 2.0 authentication and role-based access control, enhancing system security and improving authorized user access management.
C, Python, Arrays, Hash Tables, Linked Lists, Stacks, Queues, Trees
Data Cleaning & Preprocessing, Exploratory Data Analysis (EDA), Statistic Modeling, Regression
Python, Scikit-learn, NumPy, Pandas, K-Means Clustering, Supervised & Unsupervised Learning
