Deepthi Chintapalli

• Customer Identification Programs (CIP): Expertise in systems designed to verify customer identities, crucial for anti-money laundering (AML) compliance.
• Customer Due Diligence (CDD) Software: Proficient in using platforms for in-depth customer risk assessment, ongoing monitoring, and enhanced due diligence processes.
• Transaction Monitoring Systems: Skilled in deploying and managing solutions that analyze financial transactions for suspicious activities and potential fraud patterns.
• Screening Software: Experienced with tools for screening customers against watchlists, sanctions lists, and politically exposed persons (PEP) databases to identify high-risk individuals or entities.
• Blockchain Analytics: Capable of utilizing tools to trace and analyze cryptocurrency transactions for forensic investigations and compliance monitoring.
• Artificial Intelligence and Machine Learning (AI/ML): Applied knowledge in leveraging AI/ML algorithms to automate and enhance various compliance functions, such as anomaly detection and predictive analytics.
• RegTech Solutions: Familiar with a range of regulatory technology solutions that streamline compliance processes, reduce manual effort, and improve regulatory adherence.
• Regulatory Reporting Tools: Proficient in using software designed to automate and simplify the generation and submission of regulatory reports to authorities.

Social Distancing Detector using YOLO

• Engineered a real-time computer vision system using the YOLOv3 deep learning model to detect individuals and measure the distance between them in video streams.
• Integrated OpenCV and YOLO to process live camera feeds and accurately compute inter-person distances using perspective transformation.
• Implemented a proximity alert mechanism that triggers a visual or audible warning when social distancing thresholds are violated.
• Applied YOLO object detection to monitor social distancing in real-time video feeds.
• Developed an alert system to notify proximity violations, useful in public spaces like airports.

Image Forgery Detection using CNN

• Developed a Convolutional Neural Network (CNN) model to identify tampered or manipulated images using pixel-level classification techniques.
• Collected and preprocessed datasets containing both authentic and forged images, including copy-move, splicing, and resampling forgeries.
• Engineered features such as edge inconsistencies and texture anomalies to improve model sensitivity to subtle image alterations.
• Achieved high detection accuracy by experimenting with deep architectures (e.g., VGG, ResNet) and regularization techniques like dropout and batch normalization.
• Visualized model predictions using Grad-CAM to highlight forged regions, improving interpretability for end users.

Emotion Detection using CNN

• Built a deep learning model using Convolutional Neural Networks (CNN) to classify human facial expressions into emotions such as happy, sad, angry, and surprised.
• Utilized publicly available datasets like FER-2013 and performed data augmentation (rotation, flipping, zooming) to enhance model generalization.
• Preprocessed input images using grayscale conversion, face alignment, and normalization to improve training consistency and performance.
• Fine-tuned CNN architectures (e.g., custom CNN, VGG) to achieve optimal classification accuracy while minimizing overfitting.
• Integrated OpenCV for real-time face detection and emotion classification from live webcam streams.
• Evaluated model performance using confusion matrix, accuracy, and cross-entropy loss, achieving over 85% validation accuracy.

Digital Thermometer (IoT)

• Designed and implemented a digital temperature monitoring system using Arduino Uno and an LM35/DS18B20 temperature sensor.
• Programmed microcontroller logic in C/C++ to read, process, and display real-time temperature data on an LCD screen.
• Integrated the system with IoT protocols to enable remote monitoring via Wi-Fi using ESP8266 module and cloud dashboards (e.g., ThingSpeak).
• Calibrated sensor readings for higher accuracy and implemented threshold-based alerts for temperature deviations.
• Packaged the prototype into a compact and power-efficient form factor, suitable for indoor and outdoor health/environmental monitoring.
• Tested the system under varied ambient conditions to ensure robustness and reliability in real-time data logging.

• Qualified for state-level competition in Polytech Fest (2018)
• Winner at District-level Polytech Fest.

• Core member, Fem-Flare Fest, KLU
• Organizer & Coordinator, Zrotriya and Samyak Fests
• Member, PULSE - Student Body Organization
• Lead Organizer, District-level Polytech Fest