Build AI Workflows That Make You Indispensable

The most valuable thing I built was the capstone project Slackbot. Not only because I built something useful for my work, but also because it requires thinking and implementing with an end-to-end mindset. It was very useful to learn about each individual piece, but understanding and seeing how it all comes together and runs as a single workflow is most valuable to me.

I now feel more confident designing and deploying AI workflows, which will help me stay relevant and is very integral in the field and in my role going forward, especially in the next year. The most valuable part for me was building a data preprocessing agent that works. Also, learning how to do high-level AI/ML architecture for data science projects is something crucial in my new role.

I would highly recommend this course to any data professional who hasn't ventured into the AI space yet. It's non-platform specific, well-run and keeps the focus where it belongs, making it especially good for time-poor professionals who need to get up to speed quickly.

The biggest impact is that I can now engineer with LLMs, not just use them: designing semantic layers, building LangGraph state machines, connecting PostgreSQL, Gmail and Slack APIs, managing agent state, and handling failure with retry logic and self-healing loops. Over the next 6 to 12 months, I expect these skills to help me move from building AI prototypes to shipping production AI systems in a full-time role.

I feel much more confident about the space of AI tooling and workflows period. Coming in, everything seemed like impenetrable buzzwords that had no relevance to my lived data science experience. Now I am seeing how that really synthesizes.

The program gave me confidence in designing and building AI-powered applications, not just using AI tools. I expect these skills to help me contribute more effectively to AI initiatives in the enterprise. The most valuable part was building a complete conversational analytics application from scratch. My capstone project was a Slack-based analytics assistant that could answer business questions, generate tables and charts, and use guardrails to reject unsupported requests. This also taught me how to efficiently use AI as a build-partner, not just a thought-partner.

Before the program, I didn't have a clear framework for how to approach implementing AI systems. The bootcamp helped me understand how these systems should be structured and organized in real projects. I would recommend this bootcamp to anyone working in data who lacks a structured plan for implementing AI in their organization. Even if you only automate one or two processes, you gain the insight needed to grow into the role.

We're only starting to see how the data landscape is being reshaped, and the skills from this bootcamp will allow me to increase both the value and speed of my contributions. I would recommend this bootcamp to anyone in data science or analytics who feels stuck and doesn't know where to start with AI. If you want direction, want to stay ahead, and actually build tools that add real value to your workflows, this program delivers that.

The most valuable part of the program was the overall experience and guidance throughout the process. The way Andres structured the bootcamp, along with his consistent mentorship, helped me understand not just the concepts but how to actually apply them in real projects. During the program, I built several AI agents, including systems that coordinate multiple agents to work together.

A key realization was that building useful AI solutions is not just about prompting an LLM or calling an API. It's about designing the full system: how context is retrieved, how data is accessed, how reasoning is structured, and how reliable results are delivered to the user. Working on the architecture for a talk-to-your-data / agentic analytics system pushed me to think more seriously about how these components fit together in practice.

The most valuable thing I learned was building a data-cleaning agent and understanding how to design a semantic layer. That experience helped me see how AI workflows can be structured to work reliably with real data. It gave me a much clearer picture of how AI can actually be integrated into everyday data work.

The bootcamp gave me the knowledge to design AI workflows that can streamline and supercharge everyday work. I also feel much more confident navigating both current and emerging AI technologies to drive efficiency. One of the most valuable things I built during the program was a talk-to-your-data Slackbot.

I'd recommend this bootcamp to everyone who knows how to code. I can't imagine it not being useful for every worker. The way the course is structured with every project building on the one before it is helpful for implementing and reinforcing concepts.

I found a lot of value in this program. Defining an AI workflow from end to end, learning Cursor as a new IDE, understanding the value of the semantic layer, agentic systems, AGENTS.md standardization, PandasAI, and creating and debugging an end to end AI workflow solution were all extremely valuable. I would recommend this bootcamp to learn from a highly experienced professional like Andres and to obtain real work experience to improve my position as a Data Scientist.

I joined the bootcamp to get more exposure to industry-grade tools and strategies. I learned a lot including MCP integration, system design (specifically GenAI system design), semantic layers, and building the capstone Slackbot. If you want to learn new emerging tools that are widely used in the industry, this is a great course to learn from.

It is totally worth it if you want practical applications you can use in your day-to-day work, or if you want to future-proof your skills and career for the AI-driven era.

I was intrigued by the content of the course. It promised not only data science with AI but future-proof concepts like agency and semantic modeling, and it delivered on all parts. I was pleasantly surprised by learning and seeing how colleagues progressed and had different views on the way we work with LLMs. The teacher has a good understanding and knowledge in this which he also shared.

I wanted to learn how to deliver production-ready AI agents that could benefit the business by generating insights from structured and unstructured data. The most valuable part was learning and putting into practice the latest technologies and standards such as semantic layers, MCP servers, and AGENTS.md, along with development tools like Cursor AI, all through the capstone project. I'd say being hands-on with Python is a must, and being ready to learn and unlearn the way AI agents are being delivered. It's not easy, but if you use failure as a stepping stone and don't hesitate to work hard, it's the most enjoyable and doable course out there!