Building Agentic AI Applications with a Problem-First Approach
Learn how to design & build impactful AI products grounded in real experience!
Categories: Agentic AI
Limited-time: use code BAAAPCENA for 15% off · ends 05 Sept 2026
Overview
Our instructors each bring more than a decade of experience across AI research and applied AI, including publications at top conferences, patents, and building and leading more than 50 AI deployments for companies including Microsoft, Google, AWS, OpenAI, Samsung, and more! 🚨 Should you really buy an AI course (including this one)? Read this guide No hard prerequisites: Supports no-code and code formats 40+ hours of live 2-way interaction with instructors instead of passive 1-way lectures Continue learning long after the course through our weekly community and guest sessions. To estimate the outcomes of the course, see our capstone projects
What you'll learn
- Develop a Problem-First AI Intuition For Enterprise Use Cases Identify where agentic AI can add value by reframing business challenges through a systems lens Understand why traditional software assumptions fail in AI-driven environments Evaluate tradeoffs between model choices, latency, performance and cost
- Master Enterprise Level Context Engineering Identify how to incorporate tools, retrieval, memory, etc., to build autonomous systems. Learn best practices for designing agent-based products like agent harnesses, vertical AI agents, long-running agents, and more. Identify research and enterprise-backed methods to build scalable agentic products.
- Build an Evals Driven Mindset & Iterative Design Learn to make decisions tailored to business constraints, understand when & how to apply AI effectively & build a multi-agent application Implement evaluation and guardrail techniques using LLM judges and semantic scoring Learn architecture-specific evaluation mechanisms ( retrieval, tool calling, autonomous agents etc.)
- Learn Existing Trends & Design Patterns Get an insight into frontier practices at enterprises like post-training, inference optimization, using SLMs, fine-tuning, etc. Analyze multi-agent coordination patterns and challenges & learn about protocols like MCP/A2A as well as constructs like skills/plugins Identify existing enterprise trends and when to use them thoughtfully: harness engineering, loop engineering, and more!
Schedule
- 4 weeks, 4 live sessions, one each week. The first is 12 Sept 2026, 12:00 AM GMT+8, and they run weekly after that.
- Every session is recorded. You keep the recording for 30 days after each one, so you can catch up if you ever miss one live.
Curriculum
4 weeks · 4 live sessions · tap a week to see what you build
Week1Session 111 Sept 2026
Week 1: Introduction to Generative AI, Agentic AI & Building Applications
Core ideas behind AI, ML, Deep Learning, and GenAI, Agentic AI, what they are and how they differ. How AI models are trained: pre-training, fine-tuning, and RLHF. Input/Output framework to judge how useful or new a AI model really is. How to design AI solutions, deal with non-determinism, and why data matters. How to build iteratively and pick the right setup
Week2Session 218 Sept 2026
Enterprise AI Trends Workshop
Context Engg. in 2026 & Building Workflow Agents For The Enterprise
Week3Session 325 Sept 2026
Week 3: Agentic Retrieval, Search & Self Improvement
How to choose the right embeddings, vector databases and architecture for RAG Advanced methods to improve retrieval and evaluations for RAG Types of Agent memory and how to implement them
Week4Session 402 Oct 2026
Week 4: Building Autonomous Agents/Multi-Agents & Fine-Tuning
Why Memory Is the Hardest Problem in AI Agents
Projects
Project - Room: Portable AI Agents
Assignments
[Build] Assignment 1 : Setup And Testing Your Environment
Do the setup and testing
Feedback
Each week you will ship the project and I will leave a written feedback.
Capstone
Project - Digital Twins for AI-Native Enterprise Software
About Cena
Reviews
Reviews appear here once 3 learners have completed this session.
Free resource
AI Introduction
For people who wants to regain back their knowledge in AI
FAQ
Who is this cohort designed for?
This cohort is built for people who want to design and build agentic AI applications, whether or not they write code. It supports both no-code and coding approaches, so practitioners, product thinkers, and developers are all welcome.
Do I need a technical background or prior AI experience to join?
No hard prerequisites are required. The program is structured to accommodate participants across different technical levels, and instructors adapt to both code and no-code formats throughout the sessions.
What does the live learning experience actually look like?
The cohort includes more than 40 hours of live, two-way sessions where you can ask questions, get feedback, and interact directly with instructors rather than watch pre-recorded lectures. This is an active learning format, not passive consumption.
Who are the instructors and what is their background?
The instructors each have more than a decade of experience across AI research and applied AI, with publications at top conferences, patents, and involvement in over 50 AI deployments for organizations including Microsoft, Google, AWS, OpenAI, and Samsung.
What will I have to show for it at the end?
Participants complete capstone projects, which you can review on the program page to get a realistic sense of what past cohort members have built. These projects reflect the problem-first approach the curriculum is built around.
Is there any support or community after the cohort ends?
Yes, learning continues after the cohort through a weekly community and ongoing guest sessions, so you stay connected and keep building on what you learned beyond the final week.
More from Cena
Bundles, workshops, cohort programs, and Sprint Lessons from this practitioner.
