Rahul Verma

Software Engineer

• Developed custom activity for Journey Builder using Salesforce Marketing Cloud SDKs and REST APIs, enabling advanced journey functionality and custom event triggers.
• Designed and implemented customized, multi-step customer journeys in Journey Builder, integrating advanced decision splits, event triggers, and data updates to create highly personalized marketing experiences.
• Developed and optimized AMPscript-based dynamic email templates, enabling real-time content personalization by leveraging subscriber attributes, data extensions, and external data sources.
• Conducted market analysis of competitor campaigns, compiling detailed reports on strategies, content styles, and performance benchmarks to guide marketing decisions.
• Optimized complex SQL queries in Automation Studio to transform, deduplicate, and aggregate marketing data across multiple data extensions for advanced segmentation.
• Conducted in-depth performance testing and optimization of complex data flows in SFMC to enhance scalability and speed.
• Utilized Einstein STO to analyze historical engagement patterns, optimizing send times for maximum customer interaction.
• Generated detailed performance reports for campaigns, analyzing open rates, click-through rates, and unsubscribes, and shared actionable insights with the marketing team.
• Managed real-time debugging and resolution of high-priority issues during campaign execution to ensure flawless delivery.

State University of New York, Binghamton

• Operating Systems
• Programming Languages
• Design and Analysis of Algorithms
• Computer Architecture and Organization
• Introduction to Data Mining
• Database Systems
• Introduction to Computer Security
• Social Media Data Science Pipeline
• Science of Cyber Security

SQL Injection- Offense, Detection, Prevention and Deception

Developed a comprehensive security framework incorporating Offense, Detection, Prevention, and Deception mechanisms to safeguard web applications against SQL Injection (SQLI) attacks.

Engineered a deliberately vulnerable PHP-based web application to simulate real-world SQLI attack scenarios.

Designed and trained a Naïve Bayes Classifier to detect anomalously long input strings indicative of SQLI attempts, enhancing real-time threat detection.

Implemented three-layered prevention mechanisms—Prepared Statements, MySQLi real_escape_string, and Input Validation—to fortify the application against SQLI exploits.

Developed an ML-driven deception mechanism that misleads attackers while logging malicious attempts for forensic analysis.

Utilized a technology stack including PHP, MySQL, XAMPP, HTML, and CSS to build and test the system in a controlled environment.

Social Media Data Collection System, Dataset Measurements and Sentiment Analysis

Developed and deployed web scrapers to collect structured and unstructured data from social media platforms, including Twitter (sports-related content), Reddit (comments), and YouTube (comments), over a one-month period.

Designed an ETL pipeline to preprocess and store the extracted data in MongoDB, enabling efficient retrieval and analysis.

Conducted negative sentiment and toxicity analysis using Natural Language Processing (NLP) techniques to assess user engagement and content moderation trends.

Generated data-driven visualizations using Matplotlib, providing actionable insights through interactive plots and sentiment trend analysis.

Student Registration System

Developed a database-driven application to facilitate core student registration processes, including course enrollment, record management, and schedule tracking.

Designed and implemented PL/SQL-based database logic, incorporating Sequences, Stored Procedures, and Triggers to enforce business rules and maintain data integrity.

Integrated JDBC for seamless communication between the application and the relational database, ensuring efficient query execution and transaction management.

Built a text-based user interface, enabling interactive registration and administrative operations through command-line inputs.

Image Classification on Edges

Conducted standard image classification using ResNet-18 on benchmark datasets (CIFAR-10, CIFAR-100, SVHN, and Tiny ImageNet) and reported classification accuracy.

Utilized Dynamic Feature Fusion for Semantic Edge Detection (DFED) and Dense Extreme Inception Network (DexiNed) to extract and visualize edge features from images, leveraging pre-trained models provided by the authors.

Developed an edge-enhancement pipeline by combining detected edges with original images to create edge-enhanced datasets for improved feature representation.

Evaluated the impact of edge information by performing ResNet-18-based classification on (1) edge-only images and (2) edge-enhanced images, analyzing classification accuracy across all datasets.

Conducted a comparative analysis to assess the effectiveness of incorporating edge-based features in deep learning models for image classification improvements.

Twitter Trend Analysis

Published Paper (IJSRCSEIT)

Developed an automated pipeline for ingesting and processing a continuous stream of raw tweets in real time.

Implemented natural language processing (NLP) techniques to filter out irrelevant noise and extract informative tweets for further analysis.

Designed and deployed a trend detection algorithm to identify and predict emerging trending topics in their early stages.

Built an interactive GUI to structure Twitter data by converting textual content and sentiment analysis results into tabular format for better visualization.

Integrated an automated clustering feature to categorize tweets based on geolocation, time zones, and contextual similarities.

Conducted statistical analysis on tweets, generating insights through histograms, scatter plots, and bar graphs to visualize trends based on location and date.

Implemented a user search functionality to retrieve detailed user-specific information, enhancing the system's analytical capabilities.

Simple Pipeline Simulator