We’ll hire a number of strong interns at Cloudforce this summer. This program is about the one who works alongside the CEO — building AI agents on nebulaONE, the platform we’ve deployed across 100+ universities and health systems, and which earned us Microsoft’s 2025 Partner of the Year award for Education.
The internship is 8–12 weeks, flexible around your other obligations. The work is real. The customers are real. And for the right person, what happens after the summer is the part that actually matters: a full-time AI role at an organization that is pushing the boundaries of the Frontier — deep agent experience, real relationships at the institutions actually shaping how AI shows up in education and healthcare, and a track record of shipped work in the most consequential moment AI will ever have.
This isn’t just an internship. It’s a launchpad.
To find that person, we’re running No Resume Required — a 7-day public hackathon with a $5,000 stipend and the seat attached.
This page has everything you need: how it works, what to build, how you’ll be judged, the timeline, and the rules.
Registration is required. The final submission link will only be shared with registered participants before the Monday deadline.
Please register by end of day Wednesday, May 6 so we can manage volume appropriately and share information confidentially with each of you throughout the week.
👉 [Register your participation here] (it only takes 2-3 minutes)
If you’re serious about this, register today. The earlier you’re in, the earlier we can get you what you need.
Every university and health system on nebulaONE captures lectures, training videos, and recorded sessions. They sit on terabytes to petabytes of great video content. Much of it is watched once, by some of the people, sometimes — and then archived.
There’s enormous untapped value in that footage. Your job is to build a multi-agent AI system that unlocks it.
You’ll build one web application that takes public YouTube lecture URLs as input and produces something genuinely valuable for real human beings. Your app will be tested live by us, with YouTube links of our choosing, during the judging window.
This is not a “pick one scenario” challenge. It’s a layered build.
It is one application, not three. Whatever you build sits behind a single URL.
⚠️ Read this twice. A beautifully executed Capability 1 will beat a half-broken attempt at all three — every single time. The hardest decision in this build is which capabilities to attempt. We’re not rewarding ambition for its own sake. We’re rewarding judgment. Doing Capability 1 superbly will outperform doing all three sloppily. What you choose to attempt — and what you choose to leave out — is part of the test.
A college student preparing for an exam. They missed a lecture, or watched it and didn’t absorb it, or are coming back to it days later with rusty memory. They have limited time, competing obligations, and need to study in short focused bursts.
The student pastes a single lecture YouTube URL. Your app turns it into a complete personalized study environment:
Students who use a study companion like this can compress 60 minutes of passive lecture viewing into 15 minutes of active, targeted study — and retain more. For students with learning differences, language barriers, or scheduling constraints, this is the difference between passing and failing. For institutions, it’s the difference between video that gets watched once and video that compounds in value across a semester.
A faculty member who has just recorded a lecture and wants to improve it before publishing — or wants to understand what’s working and what isn’t in lectures they’ve already published. This is voluntary, private, and explicitly not surveillance: the report is for the faculty member only.
The faculty member pastes their lecture URL. Your app audits it across pedagogical, accessibility, equity, and clarity dimensions. It produces a private report only they see, with a prioritized fix list — “if you change one thing before publishing, change this” — and specific timestamped suggested rewrites.
Faculty don’t get this kind of feedback easily. Peer review is rare; student evaluations come months later, in aggregate, and are widely known to encode bias rather than instructional quality. A trusted, private, restraint-driven audit gives faculty a tool to improve their own teaching — on their own terms, before the lecture goes live to students. Institutions that adopt tools like this measurably reduce accessibility complaints and improve student outcomes in foundational courses. Faculty who use them keep using them — because they trust them.
A provost (or equivalent academic leader) who needs to know whether a course is delivering on its stated learning objectives.
Your app ingests multiple lecture URLs from a single course and produces a curriculum map that compares what was actually taught against the course’s stated learning objectives. The provost sees, at a glance: which objectives were covered, how thoroughly, and which were under-served or missing entirely.
Provosts are accountable for what their institutions promise students — in catalogs, in accreditation reviews, in marketing copy. Today, they have almost no way to verify that promise at scale. They rely on syllabi (which describe intent, not execution) and student evaluations (which don’t measure coverage). A curriculum-mapping tool gives them direct evidence — drawn from the lectures themselves — of whether a course is doing what the catalog says it does. For an institution built on trust, this would serve as a quality assurance breakthrough.
Whichever combination of capabilities you attempt, here’s what’s non-negotiable.
At least three distinct AI agents with genuinely different jobs and real orchestration — not three separate prompts in a loop. You’ll be asked to justify every agent’s existence on camera.
A single web app, behind a single URL, that:
Your app must be live and functional at a public URL from the moment you submit through Friday, May 15 at 5:00 PM ET — the close of our judging window. We will paste YouTube URLs of our choosing into your app and watch it work. If your app is down when we test it, we cannot judge it.
You may host wherever you like. A few notes:
A single diagram (PNG, PDF, PPTX, DOCX) showing your agents, your data flow, and your deployment architecture. Hand-drawn is fine. A whiteboard photo is fine. Figma, Canva, Vision Lucid — all fine. We just need to see how you thought about it. No diagram, no submission.
A single screen recording, between 4 and 8 minutes in which you:
Part A — Demo (most of the video). Walk us through your app end-to-end with a YouTube URL of your choosing. Show every capability you built. Show the failure modes you handled. Show what you’re proud of.
Part B — On camera, briefly: Answer this question: “Walk us through one tradeoff you made between scope and quality during this build. What did you cut, what did you polished, and why?”
We’re looking for product judgment — the kind a Product Manager exercises every day. The honest answer beats the polished answer. We’d rather see you stumble through a real explanation than glide through a rehearsed one.
Upload to YouTube as unlisted and include the link in your submission.
When we sit down to judge, here’s exactly what we’ll do:
In both cases, your app must hold up. Reliability is a judging criterion, not a nice-to-have. A submission that wins demos and breaks on judging videos loses to one that does less, but does it every time.
Five axes. Strong submissions show up on all five.
This program is for the person who will work on-site at our National Harbor, MD headquarters, 5 days per week, for the duration of the internship. This is a hard requirement. No exceptions. No remote. No hybrid.
If you cannot be onsite five days a week in National Harbor for a minimum of 8 weeks this summer, unfortunately this program is not for you. The $5,000 stipend goes to the person who wins the seat. This seat requires being physically in it.
We’ve built Cloudforce on the conviction that the right person, at the right moment, with the right attitude, can change the trajectory of both their own career and our client’s success. We’ve lived it. We’ve watched it happen. And we believe the AI moment we’re in right now will define careers for the next twenty years.
If you’re early in yours and you’ve been waiting for someone to actually look at what you can build — not where you went to school, not what’s on your resume, not who you know — this is that moment. And Cloudforce is that company.
This will not be easy, but nothing that’s life-changing ever is. May the force be with you and happy building. ✊
Now, if you’re up for the challenge 👉 [Register your participation here]
As fearless head-nerd and President/CTO of Cloudforce, Husein leads his eclectic team of engineers in the development of complex cloud solutions to support mission-critical systems for dozens of programs across industries and government. In addition to all things Microsoft cloud, Husein is also passionate about giving back to his Prince George’s County community and believes it is the responsibility of every business to help develop the types of communities its members and customers want to live and do business in. Husein currently serves on the board of directors of STEMship, a local non-profit aimed at creating STEM learning opportunities for under-privileged youth. Husein is also head of the Technology and Innovation Committee for the Prince George’s Chamber of Commerce.
Momentum is building at Cloudforce, and the numbers tell a compelling story. Over the past two years, the company has grown its team by 83%, more than doubling in size. Behind that growth is...
Cloudforce has been awarded funding through the Maryland Lighthouse Industries & AI Internship Program, to support the next generation of AI and cloud talent across Maryland. As a Microsoft...
NATIONAL HARBOR, Md., April 1st, 2026 — Following successful expansion across three continents, Cloudforce has announced plans to establish a world-first...
