Adam Regragui
Data & Applied AI Engineer
Engineering reliable data systems and AI solutions for real-world impact
Data & Applied AI Engineer
Engineering reliable data systems and AI solutions for real-world impact
Hi, I'm Adam, a data & AI engineer who transforms complex data challenges into elegant, production-ready solutions. From finance to advertising to regulatory tech, I've designed systems that process terabytes of data, power real-time decision-making, and deliver measurable business impact.
This portfolio isn't just a showcase. It's a window into how I approach engineering. You'll find real metrics, interactive visualizations, and the technical insights behind each solution.
Whether you're here to evaluate my work, explore my technical philosophy, or discuss how I can help solve your data challenges, I've built this site to give you a complete picture of what I bring to the table.
Let's dive in!
Scroll horizontally to explore my journey
Real metrics from production systems
Batch to Real-Time Streaming Architecture
Entity-Aware Chunking & Hybrid Retrieval
Before
After
Event Processing & Donor Profiling System
Auto-Scaling & Resource Management
Infrastructure Before
60%
Infrastructure After
36%
Ask Spoofy anything about my experience, skills, or projects
Ask me about Adam's background
Hi! I'm Spoofy, Adam's cat. Ask me about:
Full detailed CV
See how my profile matches your job requirements
Semantic analysis uses embedding models to capture contextual meaning and relationships between concepts. It recognizes that "data pipelines" and "ETL architecture" are related, even without exact word overlap, identifying transferable skills across different terminology and domains.
Keyword matching uses exact matching to identify specific technical requirements. It ensures critical skills like "Python", "Airflow", or "Snowflake" are explicitly present, catching hard requirements that semantic similarity might underweight.
The balance: Pure semantic matching may overvalue vague conceptual overlap while missing explicit must-haves. Pure keyword matching ignores context and penalizes candidates who use different terminology for the same skills. A 50/50 blend provides the most robust assessment, capturing both hard requirements and deeper skill alignment.
Feel free to adjust the balance between the two as needed.
Paste a job description and click "Analyze Match"
Click on a domain to explore
