This year, I'm thrilled to be working on a variety of full-stack projects utilizing Python, SQL and HTML, as well as diving into AI/ML innovations, all while studying at USC!

StackSense is an AI monitoring toolkit for Python that helps developers track token usage, latency, cost, and provider performance across their applications. It wraps existing AI clients with minimal integration overhead, turning invisible model spend into clear, actionable analytics through dashboards, exports, and alerts.

Key Features:

• Drop-in client monitoring with a lightweight integration for existing Python AI workflows
• Multi-provider tracking across OpenAI, Anthropic, Google, Cohere, Groq, Pinecone, ElevenLabs, and more
• Real-time analytics for usage, latency, spend, and provider-level cost breakdowns
• Developer tooling including CLI commands, dashboards, alerts, exports, and framework middleware

Tech Stack:

• Core SDK: Python 3.9+, Pydantic, Requests, Rich
• Data Layer: SQLAlchemy, SQLite, PostgreSQL
• Dashboard & CLI: Flask, Click, Plotly, Pandas
• Integrations: OpenAI, Anthropic, Gemini, Mistral, Cohere, Groq, Pinecone, ElevenLabs, Replicate
• Framework Support: FastAPI, Flask, Django
• Other: Webhooks, CSV/JSON exports, decorator API, self-hosted analytics

🔗 GitHub Repo

TRY LINGO NOW

Lingo is an AI-powered speech coaching platform that provides personalized language fluency feedback through real-time audio analysis and multimodal AI. It gamifies learning with vocabulary challenges and tracks progress using comprehensive performance analytics.

Key Features:

• Real-time speech analysis with WebRTC audio capture
• Multimodal AI processing using Google Gemini 2.5 Flash
• Gamified progress tracking to enhance speaking proficiency
• Performance analytics with visual dashboards

Tech Stack:

• Frontend: React 19.1.0, TypeScript, Tailwind CSS, Vite
• AI Processing: Google Gemini 2.5 Flash, Multimodal Analysis
• Audio Tech: WebRTC MediaRecorder, Web Audio API
• Data Visualization: Chart.js, React-ChartJS-2
• Storage & Deployment: LocalStorage, Progressive Web App
• Other: Real-time volume visualization, structured JSON schemas, mobile-first responsive design

🔗 GitHub Repo

A retrieval-augmented generation (RAG) system that can search, analyze, and summarize large datasets or web content. The agent uses vector embeddings for semantic search and provides contextual, citation-rich answers to complex queries.

Tech Stack:

• Backend Framework: Python (CLI-based application)
• AI/ML: OpenAI GPT-4 API, LangChain, OpenAI Embeddings
• Vector Database: ChromaDB (local vector storage)
• Document Processing: PyMuPDF (PDF extraction), BeautifulSoup (web scraping)
• Text Processing: LangChain RecursiveCharacterTextSplitter, tiktoken
• Output Generation: python-docx (Word documents)
• Environment Management: python-dotenv
• Web Interface: Streamlit (optional UI)
• Other: requests (HTTP client), tempfile (temporary file handling)

🔗 GitHub Repo

A lightweight, high-performance URL shortening service with click-tracking analytics, custom aliases, and rate-limiting for security. Built with a REST API backend and containerized for scalability.

Tech Stack:

• Backend: Flask (Python), REST API
• Database: PostgreSQL, Redis (caching)
• Hosting & Deployment: Docker, AWS EC2, Nginx
• Other: JWT Authentication, Rate Limiting, SQLAlchemy ORM

🔗 GitHub Repo

• Designing scalable microservices for AI applications
• Building monitoring and observability tooling for production AI systems
• Experimenting with low-latency inference in production AI systems
• Building educational technology tools with gamified learning mechanics
• Optimizing data pipelines for real-time analytics

Email: abdulkvng@gmail.com LinkedIn: Abdulrahman Sadiq 😄 Pronouns: He/Him

⚡ Fun fact: I’m a huge fan of sci-fi movies, futuristic tech, and always finding time to learn something new!

📫 How to reach me:

• Email: abdulkvng@gmail.com
• LinkedIn: https://www.linkedin.com/in/abdulrahman-sadiq/