NVIDIA AI Bootcamp

About this Course

This will be the first time Nvidia’s Deep Learning Institute (DLI) is offering this class in a public setup. The bar for what AI-powered agents can do has been steadily rising over the past few years, and new innovations allow them to not only engage in conversations but also utilize tools, conduct research, and execute on complex objectives at scale. This course empowers you to develop sophisticated agent systems that can execute on deep thought, research, software calling, and distributed operation. Throughout the course, you’ll gain hands-on experience in designing agents that efficiently retrieve and refine information, intelligently route queries, and execute tasks concurrently using orchestration tools like LangGraph and sound software engineering practices. By the end of the course, you will have a solid foundation in agent architectures and will be able to construct interesting agent-like integrations to complement your existing workflows and software stacks.

Course Details

Duration: 8 hours
Level: Technical – Intermediate
Course Prerequisites:

Introductory deep learning knowledge (including attention mechanisms and transformers). Experience from DLI’s Getting Started with Deep Learning or Fundamentals of Deep Learning is preferred.
Intermediate Python proficiency (including object-oriented programming and familiarity with ML libraries). Tutorials like Python Tutorial (w3schools.com) or equivalent practical experience suffice.

Tools, libraries, frameworks used: Python, PyTorch, HuggingFace, Transformers, LangChain, and LangGraph.

Note: We cannot accept refund or cancellation requests for this workshop. Seats are limited and all prerequisites MUST be met, so please plan accordingly.

REGISTER

Learning Objectives

By participating in this course, you will:

Understand the strengths and limitations of LLMs, and why agent-based paradigms help us to empower them in our modern software landscape.
Learn to produce structured outputs to enable machine-parseable function calls or API integrations.
Explore retrieval mechanisms and knowledge graphs for domain knowledge.
Experiment with multi-agent orchestration using frameworks like LangGraph.
Implement resilient systems and data flywheels for production-oriented deployments.

Topics Covered

We start with basic LLM usage and agent fundamentals, covering structured outputs, retrieval, and knowledge graphs. We then move to multi-agent concurrency, data flywheels, real-time constraints, and scaling considerations—finishing with a final assessment that has you interfacing with a scalable multi-tenant agent API.

Course Outline

1. Fundamentals of Agent Abstraction and LLMs

Discuss LLM capabilities & pitfalls

Introduce agents as a task decomposition abstraction.

Demonstrate minimal agent with free-text LLM calls

2. Structured Output & Basic Fulfillment Mechanisms

Bottlenecking LLMs with JSON/task-based outputs.

Ensure domain alignment & stable schema enforcement.

Introduction to cognitive architectures.

3. Retrieval Mechanisms & Environmental Tooling

Formalize environment access strategies for agents to interface with other systems.

Develop tool interfaces for external data repositories (DBs, APIs)

Use vector-RAG-coded for semantic retrieval over document sets

4. Knowledge Graphs & Document Graphs

Plan progression of data from raw docs to canonical forms.

Motivate threshold/equilibrium objectives for driving event loop.

Build state pools/ontologies for robust domain coverage

5. Multi-Agent Systems & Frameworks

Decompose tasks among specialized agents

Formalize communication buffers and process distribution schemes.

Differentiate between different frameworks and their unique approaches.

6. Data Flywheels & System Hardening

Capture usage logs, refining domain constraints, or sub-models

Implement human-in-the-loop oversight for error correction

Iterative improvement & pipeline simplification using real/synthetic data

7. Scaling & Productionalization

Discuss production-oriented considerations like resource management, concurrency, resource utilization, multi-tenancy

Motivate framework-agnostic modular deployments (meta-frameworks) and their selection criteria.

8. Final Assessment

Deploy an agent endpoint that can support multiple different interactions.

Run a distributed dialog loop across the deployed server to assess satisfaction.

9.1. [Optional] Real-Time Agents

Discuss multimodal considerations and agentic use-cases that interact with the physical world.

Explore recent advances in robotics, audio systems, and world models.

9.2. [Optional] Responsible Agents

Discuss common failure modes in software design that introduce unfairness, liability, and poor software experiences.

Consider checks-and-balances systems, standards creation, and evaluation needs.

NVIDIA AI Bootcamp

About this Course

Course Details

REGISTER

Learning Objectives

Topics Covered

Course Outline

Team Members

Kevin Lee

Marcus Manos

Justin Brandenburg

Rana Harber

Huiwen Ju

Veera Subbu Pendalaya

Find Your Ticket Match