Ali MEHIZEL


AI Engineer at AI-GRID

Greetings! I have a Master’s degree in Statistics and Data Science from ENSSEA, focused on Computational Science and Large-Scale AI. AI Engineer at AI-GRID, Full-Stack Web Developer specializing in JavaScript and Python. I’m especially interested in research contributions in deep learning optimization and computer vision.

Skills

FastAPI Django ReactJS HTML5 CSS3 PyTorch PostgreSQL SQL Docker Docker Swarm LangChain LlamaIndex Pandas NumPy Seaborn Django REST API vLLM Triton

Experience

AI GRID logo

AI Engineer

AI GRID

Nov 2025 - Present · 5 mos

Algiers, Algeria · On-site

  • Developing and scaling AI solutions for production
  • Optimizing AI models using vLLM for performance and efficiency
  • Deploying, monitoring, and maintaining cloud infrastructure
  • Collaborating with cross-functional teams to deliver AI projects
Yassir logo

Data Analyst Intern

Yassir · Internship

Mar 2025 - Jun 2025 · 4 mos

Google BigQuery, Looker (Software) and +2 skills

  • Built analytical dashboards in Looker to track product and business KPIs
  • Wrote and optimized SQL queries in Google BigQuery for reporting pipelines
  • Performed data cleaning, validation, and exploratory analysis for decision support

Featured Projects

Vidal DRF Ecommerce API

Django/DRF ecommerce backend with authentication, product management, cart workflows, search and filter, and payment support.

DjangoDRFSQLPayments
View on GitHubvisibility

Django Dashboard Expense Tracker

Dashboard application to track income and expenses with authentication, CRUD flows, chart-based insights, and export options in PDF/CSV formats.

DjangoChart.jsReportingAnalytics
View on GitHubvisibility

Bike Sharing Demand Prediction

Deep learning project for forecasting station-level bike demand using temporal modeling and data analysis for operational planning.

Deep LearningLSTMPythonForecasting
View on GitHubvisibility

Fine-Tuning LLM

Evaluation-focused LLM fine-tuning experiments comparing LoRA and baseline setups across perplexity, token accuracy, speed, and GPU usage.

LLMLoRAPyTorchEvaluation
View on GitHubvisibility

ResKAN Model

Deep image recognition architecture using Kolmogorov-Arnold Networks (KANs) to replace channel-mixing MLP blocks for improved representational power.

Computer VisionKANPyTorchResearch
View on GitHubvisibility

Contact me

Open to: research collaboration and AI/ML or full-stack engineering roles.

Timezone: Africa/Algiers (UTC+1) · Languages: English, French, Arabic

Download CV (PDF)

Let’s connect

mehizelali@gmail.com | +213674240253

school