Delivering high-quality AI-powered applications and privacy-focused solutions for diverse clients, consistently achieving top ratings and 100% on-time, on-budget project completion.
Key Achievements
- Delivered 45+ client projects leveraging Generative AI and data-driven solutions, consistently aligning with client goals and business needs
- Developed AI-powered applications and privacy-focused systems that enhanced user experience while safeguarding sensitive data
- Earned a perfect 5.0-star rating across 48 client reviews by combining technical excellence with clear, reliable communication
- Achieved 100% on-time delivery and 100% on-budget project completion, ensuring strong client trust and repeat collaborations
Technologies Used
PythonPyTorchTensorFlowHugging Face TransformersLangChainOpenAI GPT APIsStable DiffusionFastAPIDockerGit
Designing and implementing scalable data pipelines and privacy-focused systems for ad detection across 100K+ websites, integrating ML workflows, OCR, and real-time analysis to enhance user transparency and block targeted ads.
Key Achievements
- Designed and built a scalable data pipeline to support ad detection and privacy-focused analysis from 100K+ websites scraped via a custom crawler based on Tranco lists
- Engineered robust ETL processes to ingest, clean, and enrich web data using automated classification across 12 categories and 35 subcategories
- Collected and structured network-level and DOM-based signals to identify ad-related elements using community filter lists and advanced scraping methods
- Extracted and stored 'Why This Ad' metadata at scale to analyze targeting strategies, contributing to user transparency on ad personalization
- Built a semi-automated data cleaning framework to handle inconsistencies and noise, improving data quality for downstream processing and modeling
- Integrated OCR capabilities for extracting text from ad images, enriching feature sets for analysis
- Developed batch and real-time processing workflows to support machine learning pipelines for ad classification and detection
- Contributed to the creation of a Chrome extension prototype backed by engineered data flows and inference pipelines, enabling real-time blocking of targeted ads
Technologies Used
PythonApache SparkETLMachine LearningOCRWeb ScrapingCustom CrawlersData CleaningChrome ExtensionsReal-time Processing
