السلام عليكم! I’m Osama Alashkar, and my journey into AI and data started with a simple curiosity: I’ve always loved searching for patterns in sports, in markets, and in the everyday decisions people make. I realized early on that behind what looks random, there’s usually data telling the real story.
That curiosity pulled me into machine learning and deep learning, where I began teaching machines to uncover those patterns. What started as exploration turned into building end-to-end AI solutions that don’t just sit in theory, but create visible impact in the real world.
Today, I combine two passions working as an AI Engineer, designing and deploying intelligent systems, and diving into data analysis, where I clean, organize, and interpret raw data to extract insights that drive smarter decisions. For me, AI is not just about technology it’s about solving real problems with creativity, precision, and curiosity.
Most businesses struggle when messy, inconsistent, or incomplete data holds them back broken dashboards, duplicated records, missed insights, and wasted hours spent cleaning spreadsheets instead of making decisions. That’s where I step in fixing the foundation fast by cleaning, standardizing, and stabilizing data so every number can be trusted.
My unique edge in data analysis is not just reporting on the past, but forecasting and predicting how data can shape future decisions. I specialize in transforming raw information into actionable insights that drive growth, efficiency, and smarter strategies. Beyond dashboards and reports, I focus on building predictive and prescriptive models that anticipate trends, optimize operations, and uncover hidden opportunities. By blending AI expertise with a deep understanding of business needs, I help organizations move from reactive decision-making to proactive, data-driven strategies.
New Mansoura University
New Mansoura, Egypt
Professional data solutions tailored to your business needs
Transform your raw data into actionable insights with comprehensive analysis and stunning visualizations using Power BI, Tableau, and Excel.
Build predictive models and AI-powered solutions to automate processes and improve decision-making for your business.
Design and implement robust data pipelines and ETL processes to ensure clean, reliable, and accessible data infrastructure.
A showcase of my data analysis and visualization projects
A code editor that swaps one AI assistant for a small team of AI agents. They split the work, hand tasks to each other, and run the tests before moving on.
Most AI coding tools give you one assistant. It reads your prompt and the code around it, then answers. That is fine for a single function, but it starts to break down on a bigger change that touches many files, because one model has to hold the whole task in its head at once. As its memory fills up, its later thinking gets worse.
CSphere goes a different way. It is built on top of VS Code, so you keep the editor, the extensions, and the terminal you already use. Underneath, an orchestrator reads the request and hands out small pieces to focused agents. Each agent works on its own narrow slice, keeps a small view of the code, and leaves short notes for the others on a shared board. Before any step that a later step depends on, a test agent runs the project's own tests, so a broken change gets caught instead of passed down the line. Because each agent stays small, cheaper models that run on your own machine can do most of the work, and you only pay for the calls that actually run.
An orchestrator plans the job and spawns focused agents that pass work straight to each other. There is no middle manager deciding who talks next, so no extra call is wasted on routing.
When a step matters for what comes next, a test agent runs the real test suite in the terminal. If it fails, the work goes back for a fix rather than forward as a false success.
The editor indexes and searches your code locally with an on-device model. Your source and its search index never leave the device, so nothing gets uploaded to index your repo.
Billing is in credits that match the real cost of each model call. Local models cost nothing, so routine work is free and you can keep going even when your credits run out.
A canvas draws the agents as a node graph and updates as they work. You can watch each one think, call a tool, hand off, and step in when it asks a question.
A built-in gallery lets you install skills, tool servers, and packages to teach the agents new tricks. Staff review anything before it goes public.
Most AI coding tools run one model in a single loop, send your code to a server, and charge a flat monthly seat. CSphere makes a few different calls that matter:
1,000 credits a month and the free hosted models. Local models are always free.
25,000 credits a month, plus the mid-tier hosted models on top of the free ones.
80,000 credits a month and every model, including the top tier. Introductory price.
DeepLearning.AI
2024DEPI Certified
2025NVIDIA
2025DeepLearning.AI
2026I enjoy teaming up on projects where I can bring real value. So if you've got one, reach out. My inbox is always open, whether it's for a new project, a challenge to solve, or just a good tech chat over coffee. And if you feel like arguing why Barcelona is better than Madrid, go ahead I'll be waiting 😉 (proud Madridista here).