We Just Watched the Future Get Announced. Now We're Building It.
Honestly? Watching the announcements coming out of Google Cloud Next '26 felt like someone finally put words to the thing we've been trying to build for months.
Agentic technology is revolutionizing how we work and the transformation is accelerating. Google And sitting here as two second-year students in Kenya, we aren't just watching that transformation happen from the sidelines. We are trying to bring it somewhere it hasn't fully arrived yet: the university student's daily life.
The conversation at Next '26 isn't about whether AI can do things anymore. It's about building systems where AI actively works, decomposing goals, coordinating tasks, delivering real outcomes. When we heard that framing, we looked at each other and thought, that's exactly what students need. Not another chatbot. Not another content platform. An agentic system that works for you while you're busy being a student.
That's Octal Foundry.
The Hidden Hours Problem
Here's something nobody really talks about.
The average university student has somewhere between 2 and 4 hours of fragmented free time every single day. Time between lectures. The gap before your next class. The hour after dinner before you convince yourself to sleep. It's not a long stretch of focused time. It's scattered, unpredictable, and easy to lose to your phone.
But here's the thing: those hidden hours, added up across a semester, represent an enormous amount of untapped potential. That time could be career capital. It could be real, progressive skill-building toward a specific goal. Instead, for most students, it just disappears.
That's not a laziness problem. It's a structure problem.
When you have 45 minutes between classes and you want to learn something useful, the barrier isn't motivation. The barrier is that opening YouTube or Google means facing an ocean of content with no clear starting point. By the time you've figured out what to watch next, your 45 minutes is gone. Tutorial hell doesn't just trap people in long sessions. It kills the small windows too.
Octal Foundry is built specifically for those hidden hours. We call it what it is: a time-reclamation engine.
What We're Actually Building
Octal Foundry is an Agentic Learning Incubator. And the feature that sits at the heart of the MVP is something we think changes everything.
A student uploads their semester timetable.
That's the starting point. Not a quiz. Not a lengthy onboarding form. Just the timetable they already have. Our AI analyzes it, reads the existing structure of their academic week, and identifies the white spaces. The gaps between lectures. The free periods. The windows that currently exist but aren't being used for anything intentional.
That's where the work begins.
Once the white spaces are identified, the student describes their future. Where they want to go. What kind of career they're building toward. What they want to be capable of doing that they can't do yet. From that input, Octal Foundry's agents generate a personalized roadmap and do something we think is genuinely exciting: they give the student back an updated timetable.
Not a replacement timetable. A hybrid one.
The original academic schedule stays exactly as it is. But now the white spaces are filled. Every gap has a purpose. Every free hour has a lesson. The student gets a single document that shows them their full week, their university commitments and their personal learning journey, sitting side by side, already calibrated to fit together.
This is what it looks like in practice.
A Computer Science student wants to learn data science. They upload their timetable. Octal Foundry finds a one-hour gap on Tuesday afternoon and slots in a focused lesson on Python variables. Not a playlist. Not a course to browse through. One concept, one hour, structured to be completable in exactly the time available. The next gap gets the next concept. The roadmap moves forward one white space at a time.
Now think about a medical student who wants to align their career with AI. Maybe they've been reading about what's happening in healthcare and they want to understand where they fit in that future. They upload their timetable, they describe that vision, and Octal Foundry finds their available windows and builds toward it. A one-hour gap becomes a focused session on AI ethics in medicine. Not a vague suggestion to "look into it." A structured, scoped lesson designed to fit the time they actually have and move them meaningfully forward on a topic that matters for the career they want.
In the agentic era, a single intent triggers a chain reaction where a primary agent decomposes goals into tasks for specialised sub-agents that collaborate to deliver real outcomes. Google Cloud That is the architecture Google is scaling across enterprise workflows right now. Octal Foundry takes that same logic and points it at the student's week, filling it with purpose, one white space at a time.
No More Searching. The Right Video Finds You.
Here is one of the things we are most excited about in the MVP.
We are integrating YouTube as our primary content source, powered by LangChain. This is a deliberate choice and it matters more than it might sound.
Every student already knows YouTube exists. The problem has never been the content. The problem is the search. The decision fatigue of typing a query, scanning thumbnails, reading descriptions, clicking the wrong video, going back, trying again, and losing 20 minutes before a single minute of actual learning happens. That process is exhausting and it eats the hidden hours before the student even starts.
Octal Foundry removes that entirely.
When a student opens their hybrid timetable and sees that Tuesday's free hour is dedicated to Python variables, the platform doesn't point them toward a search bar. It surfaces the specific YouTube video that matches that lesson, calibrated to the duration of the window they have. LangChain enables us to intelligently match the concept being learned to the most relevant, appropriately scoped video available. The student opens the app, sees what they're learning, sees the video, and starts. That's it.
No searching. No second-guessing. No wasted time deciding.
The content was always out there. Octal Foundry makes sure the right piece of it is waiting for the student at exactly the right moment.
Learning That Has to Be Earned
Watching a video is not the same as learning something. We know this because we've done it ourselves, nodded along to a tutorial, felt like we understood, and then sat down to build something and realized we had retained almost nothing.
That's why Octal Foundry builds accountability directly into the structure.
Every week comes with a task. Not a suggestion. A task. For tech students, this might be a Python challenge, a small project, a function to write and test. When the student is ready to submit, they paste their GitHub URL. The platform receives it, verifies the submission, and unlocks Week 2. No submission, no progression. The learning path stays locked until the work is done.
This isn't about being harsh. It's about being honest. Real learning requires doing, and a system that lets students passively move through content without demonstrating understanding isn't actually helping them. Octal Foundry is designed to produce students who can actually build things, not students who have watched a lot of videos about building things.
The weekly task structure also means that by the time a student reaches the end of the semester, they haven't just followed a roadmap. They have a body of work. Real submissions. Real projects. Things they built during the hidden hours that used to disappear.
And at the end of it all, there is a capstone project. A final challenge that brings together everything the student has learned across the semester. Complete it and earn a certificate. Not a participation certificate. A certificate that represents a working project, submitted evidence, and a full semester of structured, intentional learning.
The holidays are not a break from this. They are dedicated project time, a longer runway for deeper work that the semester's fragmented hours couldn't fully accommodate.
The Future of Octal Foundry
What we are building for the MVP is the foundation. But the vision goes further and we want to be honest about where we intend to take this.
The long-term future of Octal Foundry has three pillars that we are genuinely excited about.
The first is an African-based AI model. Right now we are building on top of existing infrastructure, which is the right call for an MVP. But our vision is to eventually build and train our own model, one that is rooted in the African context, understands the realities of students in emerging markets, and reflects the problems and opportunities that are most relevant to this continent. This isn't a distant dream. It's a deliberate direction.
The second is African-based projects. The tasks and capstone projects that students complete should not just be technically rigorous. They should be contextually relevant. We want the projects given to students to be grounded in African problems, African datasets, African realities. A student building a solution on Octal Foundry should be building something that connects to the world they actually live in and want to improve.
The third is smart contracts for verified achievement. This is where things get really interesting. We are exploring the integration of smart contracts into the submission and certification layer of Octal Foundry. The idea is that a student must submit a genuinely working solution to unlock the next week's content and earn their weekly badge. Not a placeholder. Not a half-finished attempt. A working solution, verified on-chain, that cannot be faked or skipped. The weekly badge becomes a piece of verifiable proof. The semester certificate becomes something with real, tamper-proof credibility. In a world where credentials are increasingly easy to inflate, we want Octal Foundry certificates to mean something concrete and provable.
The agentic logic at the core of this is intentionally domain-agnostic. We are starting with Computer Science and Mathematics students because they are our beta group and they give us sharp, immediate technical feedback. But the infrastructure doesn't stop there. The medical student example isn't hypothetical. Any high-rigor field where structured, progressive knowledge acquisition matters and where students are currently left to figure it out alone is a candidate for this system.
We aren't building an app. We are building infrastructure for structured knowledge. And the more we build it, the more convinced we become that this is genuinely a revolution.
How It Started: An Idea That Found Its People.
I started Octal Foundry as a personal initiative. As a 2nd-year Mathematics and Computer Science student at Maseno University, personally stuck in tutorial hell and watching the same happen to peers, the question became impossible to ignore: why doesn't an intelligent system exist that can just take all of this content and turn it into a structured path tailored to me?
The idea existed. What it needed was the right environment to grow.
That environment was the Guild Code Community.
Guild Code is not just another online group. It is a space where talented developers find each other, where real mentorship actually happens, and where people hold each other accountable to ship things that matter. The community is led by Mr. Kelvin Obote, a product designer, AI and ML researcher, developer and enthusiast who has built Guild Code into something genuinely rare: a place where ambition meets direction. Keeping the community alive and making everyone feel genuinely supported is Madam Ivy, the community manager whose presence has made a real difference to this team at every step.
Joining Guild Code didn't just give us a space. It accelerated everything.
It was inside that community that I met Grace Karume, a 2nd-year Mathematics and Computer Science student from JKUAT. When Grace heard the idea, she didn't just understand it. She immediately started thinking about how to build it. She joined as co-founder and Lead Developer, and from that moment the initiative stopped being a personal frustration and started becoming a real product.
The support from Mr. Kelvin Obote and Madam Ivy has been consistent, practical and genuinely encouraging. Having mentors who push you, believe in what you're building and show up when it matters is something a lot of student builders never get. We don't take that lightly.
The Team
Danson Githuka leads research and agentic architecture, focused on the logic layer: how a student's goal gets decomposed, how content gets mapped to skills, how white spaces get matched to the right lesson, how YouTube videos get surfaced through LangChain, and how the entire learning path gets sequenced into something a real person can actually follow in the time they have.
Grace Karume leads technical development, owning the architecture, the build process, and making sure everything the team ships is robust and scalable.
Together, Danson and Grace lead a distributed team of developers from both Maseno University and JKUAT. Two campuses. One team. All building the tools they wish they'd had when they started.
Why the Agentic Era Belongs to Students Too
Here's what strikes us about everything coming out of Next '26.
Nearly 75% of Google Cloud customers are now using AI products to power their businesses, and the agentic enterprise transformation is accelerating across every industry. Google The infrastructure for autonomous, goal-driven AI systems is being built at extraordinary scale. And almost all of the conversation is pointed at enterprises.
We think that's only half the story.
The same principles that make agentic systems powerful for businesses, breaking down complex goals, coordinating specialised agents, delivering structured outcomes, are exactly what students in emerging markets need. The student with ambition and internet access but no mentorship, no structured curriculum, and only 45 minutes between classes is not a niche use case. That student is everywhere. That student is us.
The pace of technological change since last year's Cloud Next has never been faster. Google That's as true for two second-year students in Kenya as it is for any enterprise customer. The tools that make this buildable now simply didn't exist two years ago. We are building Octal Foundry at exactly the right moment.
The Agentic Era isn't just a corporate trend. It's an infrastructure shift. And we intend to route some of that infrastructure toward the people who need it most.
Three Months to MVP
That's the target. Three months to put something real into the hands of real students.
Every decision the team makes comes back to one question: does this help a student convert their hidden hours into career capital faster? If yes, we build it. If not, we cut it.
We are those students. We know what it feels like to have the drive and not the structure. That's not a sad story. It's the clearest possible brief for a product.
Follow the Journey
If you're a student developer who has lost hours to tutorial hell and scattered content, this is being built for you.
If you're a medical student who wants to understand where AI fits in your future but doesn't know where to start, this is being built for you too.
If you believe the Agentic Era should reach further than the enterprise, follow what Octal Foundry is building.
We are two second-year students, a talented distributed team, a community that believed in us before we had anything to show, and a vision we are completely serious about.
The hidden hours are waiting. The white spaces are there. The right video is already on YouTube. The task is already written. The roadmap is already built.
All that's left is for the student to begin.
Let's build.
Top comments (0)