Ali MEHIZEL

AI Engineer at AI-GRID

AI Engineer specializing in production LLMs, agentic systems, and speech AI. Master’s graduate in Statistics and Data Science from ENSSEA, with a focus on computational science and large-scale AI. At AI-GRID, I build and scale production AI systems—from DeepAgent architectures to DevOps with Kubernetes and Docker—delivering reliable, high-throughput inference. Also a full-stack developer in Python and JavaScript, with research interests in deep learning optimization and computer vision.

Get in touch Download CV
Ali MEHIZEL working on a laptop
Stack

Skills

FastAPI
Django
React
HTML5
CSS3
PyTorch
PostgreSQL
MySQL
Docker
Kubernetes
LangChain
LlamaIndex
Pandas
NumPy
Matplotlib
DRF
vLLM
Triton
Career

Experience

Production AI engineering and data analytics across industry teams.

AI GRID logo

AI Engineer

AI GRID

Nov 2025 – Present · 9 mos Algiers · On-site
  • Developing and scaling AI solutions for production, including vLLM optimization for performance and efficiency
  • Built production DevOps expertise across Kubernetes orchestration, Docker-based scaling, reverse proxies, and cloud infrastructure
  • Building data engineering pipelines for monitoring and observability
  • Designing monitoring pipelines for agentic AI systems
  • Working on DeepAgent architecture, agentic AI systems, and speech AI
  • Collaborating with cross-functional teams to deliver AI projects
Yassir logo

Data Analyst Intern

Yassir · Internship

Mar 2025 – Jun 2025 · 4 mos BigQuery · Looker
  • 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
Selected work

Featured Projects

Industrial systems, agentic AI, and research-driven deep learning.

Industrial Operations

Industrial ERP Platform

Scalable ERP for production, inventory, and procurement. Airflow and PySpark orchestrate data pipelines and automate industrial workflows.

AirflowPySparkFastAPIReact.jsPostgreSQLDockerKubernetesNginx
Agentic AI · RAG

Customer Support Agent

RAG-powered support agent with LangGraph multi-step workflows and Langfuse for tracing, observability, and evaluation.

LangGraphLangfuseRAGLangChainPythonLLMs
Deep Learning

Bike Sharing Demand Prediction

Forecasts station-level bike demand with temporal modeling for operational planning.

Deep LearningLSTMPythonForecasting
View on GitHubvisibility
LLM Fine-Tuning

Fine-Tuning LLM

LoRA vs baseline experiments across perplexity, token accuracy, speed, and GPU usage.

LLMLoRAPyTorchEvaluation
View on GitHubvisibility
Computer Vision · Research

ResKAN Model

Image recognition with Kolmogorov-Arnold Networks replacing channel-mixing MLP blocks.

Computer VisionKANPyTorchResearch
View on GitHubvisibility
Beyond work

Hobbies

Connect

Contact me

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

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 · +213 674 240 253