The evaluation + deployment axis is what most agent courses skip, so it's great you're putting it at the center. In my own research the part that keeps biting production teams isn't task success rate; it's trajectory-level evaluation, specifically whether each tool call was the right choice given the state the agent was in, not just whether the final answer was correct. Decomposing runs into atomic claims at the step level is how I ended up making these failures visible. One request: when you get to the evaluation week, consider framing MCP and A2A from the eval angle too, since the interfaces between services are where silent drift tends to live. Really looking forward to this.
I currently work as a data scientist at a Colombian startup (with over a year of experience), but I want to take the next step and get a remote job in the United States or another country with better career opportunities.
My bachelor degree wasn't in engineering but in business, so despite being a data scientist in title I'm constantly looking for ways to bridge that gap by acquiring the most relevant skills in the market. I have some knowledge of AI since my company uses it for classification, sentiment analysis, and so on.
My expectations for this training are to have projects I can showcase in my portfolio, as well as acquire the skills that will bring me closer to my goal of working abroad (a former colleague achieved this, so why can't I?). Do you think my expectations are realistic?
@Miguel Otero Pedrido topics are solid for sure. Also for me me it's less about whether the topics interest me and more about whether theyll still be relevant or about to explode in the next few months.
Agents, mcp, rag, A2A… I'm all for it, as long as the course is tracking what's actually being hired for and built in production right now (or is about to be).
Actually curious Miguel, do you have any predictions on what's about to blow up next in the industry? Not from a FOMO angle, but genuinely wondering if the course could be positioned to catch the next wave just as it's taking off. If someone had built solid RAG experience right before it exploded in the job market, that would've immediately stood out on a CV to any AI startup thats hiring ,right? That's the kind of timing that separates candidates i guess. So yeah, if you've got any read on what's about to be the next must have or skill thats about to be in demand , would love to see the course lean into that. Just your take on it would be nice to hear even if you think it could be an airball.
You're right that timing matters. And here's what I see: the industry has figured out how to BUILD agents (there's a new framework every week). What almost nobody knows how to do is ship them to production and make sure they actually work reliably. That's the gap I see.
Think about it … right now, most "agent" projects in companies are demos. Cool notebooks, impressive in a meeting, but they don't run in production because nobody evaluated them properly, nobody deployed them with real infrastructure, and nobody built the pipeline to catch regressions before users do.
That's exactly what this course teaches. Not "here's another agent framework", but: how do you build a multi-agent system, deploy it to real cloud infrastructure (Cloud Run, event-driven pipelines, etc.), connect it with production protocols (MCP, A2A), and — most importantly — evaluate the hell out of it before and after it ships.
If I had to name the skill that's about to explode, it's this: Agent Evaluation + Production Deployment. Some people call it AgentOps. I call it "actually shipping agents that work" xD
Every company experimenting with agents right now will need people who can take those experiments to production. And right now, almost nobody can.
So yeah, if you want to be the person who, 6 months from now, can say "I've built, evaluated, and deployed a multi-agent system on GCP, here's the live URL", that's what we're building together 💪
The evaluation + deployment axis is what most agent courses skip, so it's great you're putting it at the center. In my own research the part that keeps biting production teams isn't task success rate; it's trajectory-level evaluation, specifically whether each tool call was the right choice given the state the agent was in, not just whether the final answer was correct. Decomposing runs into atomic claims at the step level is how I ended up making these failures visible. One request: when you get to the evaluation week, consider framing MCP and A2A from the eval angle too, since the interfaces between services are where silent drift tends to live. Really looking forward to this.
I currently work as a data scientist at a Colombian startup (with over a year of experience), but I want to take the next step and get a remote job in the United States or another country with better career opportunities.
My bachelor degree wasn't in engineering but in business, so despite being a data scientist in title I'm constantly looking for ways to bridge that gap by acquiring the most relevant skills in the market. I have some knowledge of AI since my company uses it for classification, sentiment analysis, and so on.
My expectations for this training are to have projects I can showcase in my portfolio, as well as acquire the skills that will bring me closer to my goal of working abroad (a former colleague achieved this, so why can't I?). Do you think my expectations are realistic?
yes, that’s exactly what you’ll get! We’ll combine both foundations and engineering and build a project you can showcase on your portfolio!
Hi what else will you be teatching in the cohort other than agents and mcp?
hey! We'll cover RAG, Agents, MCP, A2A, Evaluation, and, more importantly, how to deploy everything to GCP (Cloud Run, GCS, Cloud build, PubSub, etc)
wdyt @Sulayam ? Do you like the topics? Would like to know your opinion man!
@Miguel Otero Pedrido topics are solid for sure. Also for me me it's less about whether the topics interest me and more about whether theyll still be relevant or about to explode in the next few months.
Agents, mcp, rag, A2A… I'm all for it, as long as the course is tracking what's actually being hired for and built in production right now (or is about to be).
Actually curious Miguel, do you have any predictions on what's about to blow up next in the industry? Not from a FOMO angle, but genuinely wondering if the course could be positioned to catch the next wave just as it's taking off. If someone had built solid RAG experience right before it exploded in the job market, that would've immediately stood out on a CV to any AI startup thats hiring ,right? That's the kind of timing that separates candidates i guess. So yeah, if you've got any read on what's about to be the next must have or skill thats about to be in demand , would love to see the course lean into that. Just your take on it would be nice to hear even if you think it could be an airball.
I’ll give you my honest take.
You're right that timing matters. And here's what I see: the industry has figured out how to BUILD agents (there's a new framework every week). What almost nobody knows how to do is ship them to production and make sure they actually work reliably. That's the gap I see.
Think about it … right now, most "agent" projects in companies are demos. Cool notebooks, impressive in a meeting, but they don't run in production because nobody evaluated them properly, nobody deployed them with real infrastructure, and nobody built the pipeline to catch regressions before users do.
That's exactly what this course teaches. Not "here's another agent framework", but: how do you build a multi-agent system, deploy it to real cloud infrastructure (Cloud Run, event-driven pipelines, etc.), connect it with production protocols (MCP, A2A), and — most importantly — evaluate the hell out of it before and after it ships.
If I had to name the skill that's about to explode, it's this: Agent Evaluation + Production Deployment. Some people call it AgentOps. I call it "actually shipping agents that work" xD
Every company experimenting with agents right now will need people who can take those experiments to production. And right now, almost nobody can.
So yeah, if you want to be the person who, 6 months from now, can say "I've built, evaluated, and deployed a multi-agent system on GCP, here's the live URL", that's what we're building together 💪
Alright buddy, tysm for taking the time. Lets go get it!