Instructor-led training · Edition 2026

Applied Artificial Intelligence

Six standalone courses, ordered from foundations to specialization — Python and machine learning, applied agricultural AI, three NVIDIA-certifying deep learning and generative AI courses, and agentic software engineering. Taught by an NVIDIA-certified instructor who builds these systems in production.

6Standalone courses
3NVIDIA certificates
8+Years in production AI
The instructor

Who teaches these programs

Hedi Fkih, PhD

DATA & AI Tech Lead with 8+ years of experience designing, developing, and deploying production-grade Artificial Intelligence solutions. PhD in Computer Science specializing in Deep Learning, Computer Vision, and Generative AI. NVIDIA Instructor & Platinum Ambassador.

Experienced in leading multidisciplinary AI teams, architecting enterprise-scale AI systems, and delivering innovative R&D projects from research to production across Generative AI, LLMs, Agentic AI, Computer Vision, MLOps, and AI-driven observability.

Strong expertise in integrating AI capabilities with Application Performance Monitoring (APM) and Digital Experience Monitoring (DEM) platforms to enhance system reliability, automate root cause analysis, and improve end-user experience. Skilled in designing intelligent monitoring solutions, leveraging LLMs, agentic workflows, and machine learning techniques for anomaly detection, incident analysis, and operational optimization.

DATA & AI Team Lead PhD, Highest Honors NVIDIA Platinum Ambassador NVIDIA Certified Instructor AWS Solutions Architect English · French · Arabic
The catalog

Six courses, foundations first

Every course stands alone and can be booked on its own. Click any course to open its full day-by-day program.

P-01 Foundation 3 days · 18 h

Advanced Python & Machine Learning Foundations

The entry point of the catalog. Builds the Python and machine learning base the five other courses assume.

+

Day 1 — Advanced Python

  • Python refresher: types, control flow, functions
  • Comprehensions and generators
  • Decorators and context managers
  • Object-oriented programming for data work
  • Modules, packaging, virtual environments

Day 2 — Data manipulation & visualization

  • NumPy arrays and vectorized computation
  • Pandas: loading, cleaning, joins, group-by
  • Visualization with matplotlib and seaborn
  • Supervised learning: regression and classification
  • Train/validation/test splits, cross-validation

Day 3 — Applied machine learning

  • Scikit-learn pipelines and transformers
  • Hyperparameter search and model selection
  • Evaluation metrics for each problem type
  • Data leakage, overfitting, reproducibility
  • Closing capstone exercise
Prerequisitebasic Python
Certificationcertificate of completion
TechnologiesNumPy, Pandas, scikit-learn
P-02 Foundation 2 days · 12 h

Artificial Intelligence for Agriculture 4.0

Built on real agricultural data — yield records, satellite imagery, disease image sets — not generic examples.

+

Program

  • AI applications in agriculture
  • ML workflow: from data to model
  • Exploratory data analysis (EDA)

Program (cont.)

  • Feature engineering for agriculture Vegetation indices: NDVI, SAVI, EVI
    Temporal features: DOY, GDD, cumulative rainfall
  • Applied ML models for classification, clustering and forecasting
Prerequisitebasic Python
Certificationcertificate of completion
Audienceagronomists, agri-engineers, agritech teams
P-03 NVIDIA DLI 2 days · 12 h

Fundamentals of Deep Learning

How a network learns, how convolutions see, and how a trained model reaches deployment — for engineers with Python and no prior deep learning background.

+

Day 1

  • Introduction to Deep Learning
  • The mechanics of a Neural Network
  • Convolutional Neural Networks
  • Image Classifier

Day 2

  • Data Augmentation & Deployment
  • Pre-trained Models & Transfer Learning
  • Advanced Architectures
  • Wrap-up & Assessment
PrerequisitePython
Certificationofficial NVIDIA DLI certificate
TechnologiesPyTorch, Pandas
P-04 NVIDIA DLI 2 days · 12 h

Building Transformer-Based NLP Applications

The transformer end to end — architecture, pre-training, fine-tuning, and the optimization work that makes a language model affordable to serve.

+

Day 1

  • Machine Learning in NLP (Text Representation, RNNs)
  • Transformer Architecture & Training a BERT model
  • Self-Supervision & Finetuning using NeMo
  • Prompt Engineering & Prompt Learning

Day 2

  • Parameter Efficient Fine-Tuning (PEFT) & LoRA
  • Post-Training Optimization, Product Quantization & Knowledge Distillation
  • Fine-tuned LLMs · Model Code Efficiency, Parallelism & Triton Inference Server
  • Wrap-up & Assessment
PrerequisiteP-03 or equivalent, PyTorch
Certificationofficial NVIDIA DLI certificate
P-05 NVIDIA DLI 2 days · 12 h

Generative AI with Diffusion Models

Generative models built from the ground up — first adversarial, then diffusion — finishing with a working text-to-image pipeline participants wrote themselves.

+

Day 1

  • Understanding Generative AI, GANs & the U-Net architecture
  • Denoising Diffusion Probabilistic Models
  • Optimizing Neural Networks for Generative AI
  • Development of an Img2Img generation model

Day 2

  • Classifier-Free Diffusion Guidance
  • Contrastive Language–Image Pretraining (CLIP)
  • Development of a Text2Img generation model
  • Wrap-up & Assessment
PrerequisiteP-03 or equivalent, PyTorch
Certificationofficial NVIDIA DLI certificate
P-06 Workshop 3 days · 21 h

Agentic Development with Claude Code & GitHub Copilot

Moves a team from ad-hoc prompting to an engineered, observable AI-augmented SDLC — six labs on real repositories.

+

Day 1 — Foundations and integration

  • LLM landscape: Claude, GPT-4o, Gemini, Copilot
  • Context engineering versus prompt engineering
  • The CDLC loop — define, structure, iterate, validate, capitalize, deploy
  • Memory files (CLAUDE.md) and instruction files (SKILL.md)
  • Lab 1 & 2 — a working agentic environment and your first project context Deliverable: an operational environment tested against a first agentic prompt

Day 2 — The AI-augmented SDLC

  • MCP: connecting agents to Jira, GitHub, GCP and databases
  • Hooks — pre-commit, post-test, on-error self-correction
  • Forking conversations for competing implementations
  • Advisor and Plan modes for full supervision
  • Lab 3 — full SDLC workflow, from ticket to Cloud Run Deliverable: a full-stack app with a working CI/CD pipeline

Day 3 — Orchestration & multi-agent systems

  • Manual orchestration — plan, validate, execute — for critical work
  • Automatic orchestration — a lead agent delegating to sub-agents
  • Lab 4 — audits with AI agents: architecture, code quality, security
  • Lab 5 — multiple agents with Git worktrees and AI-assisted merge
  • Lab 6 — agent observability: structured logs, traceability, token/duration metrics, anomaly alerts
PrerequisiteGit, CI/CD basics, one backend language
Certificationcertificate of completion
ToolingClaude Code, GitHub Copilot, MCP
Practical information

Delivery & logistics

Formats

  • On-site at your premises
  • Virtual instructor-led, live
  • Hybrid

Languages

English, French or Arabic. Materials supplied in English by default, French on request.

Materials

  • Full slide decks
  • Annotated code and lab repositories
  • Datasets used in session
  • Certificates on completion