I submitted an example of an AR storytelling integration and explained how it used interactive 3D models to teach environmental concepts.

I submitted five AR lens ideas and a digital mockup of my chosen concept for the Snap AR lens project.

I built a Moodboard Generator agent workflow in n8n and generated initial visual moodboards from text prompts.

I delivered a complete Trend Discovery Agent combining RSS, Google Search, and Amazon scraping workflows.

I delivered a competitor monitoring agent and a written competitive intelligence report that recommended tactical product and feature moves for Wayfair's rugs catalog.

I refined an AI Insights & Content Agent to turn trend and competitor data into brand-aligned blog ideas, social captions, and campaign copy for Wayfair-style marketing.

I wrote a professional STRIDE-based threat model for The Grand Marina's HYDROLOGIC water system, documenting key assets, risks, and prioritized threats.

I built data processing scripts and a mock sensor generator while practicing Python, Pandas, and JSON handling in Google Colab.

I mapped a hierarchical MQTT topic structure for a hotel's water monitoring system and reflected on what an unauthorized subscriber could see and do.

I ran four security experiments on the Grand Marina MQTT pipeline, added TLS, and produced an executive-style assessment for the hotel GM.

I implemented and tested mutual TLS controls, ran attack simulations, and documented device provisioning for a secure MQTT pipeline.

I wrote a Defense Comparison Report summarizing tests of timestamp checks, sequence counters, and HMAC signing against replay attacks.

I summarized how large language models learn and generate text, explaining pretraining, transformers, tokenization, attention, and fine‑tuning in plain terms.

I built and tested image preprocessing and OCR-ready pipelines in Google Colab, applying resizing, denoising, thresholding, and contrast adjustments to improve document text extraction.

I extracted and structured data from SDF PDFs using Python, producing a Colab notebook with the parsing pipeline.

I compared Tesseract, PaddleOCR, and EasyOCR on a scanned pharmaceutical PDF and documented differences in detection, confidence, and layout handling.

I built a Retrieval-Augmented Generation pipeline in Google Colab using LlamaIndex, implementing embeddings, chunking, hybrid retrieval, and reranking to return relevant document passages.

I compared three RAG retriever configurations on a sample SDF document and logged chunk retrievals, best excerpts, confidence scores, and a short analysis to identify the most effective settings.

I built a segmentation and classification pipeline to split bundled pharmaceutical PDFs into page-level documents and tag each chunk with metadata for routing.

I built a Gradio-based chatbot interface and updated it to extract PDF text, offer multiple response modes, and let users save chat history.