Remote proof has to travel
Remote AI work rewards artifacts that can be judged without a live narrator. A project page, repo, demo video, README, architecture diagram, and tradeoff note let a hiring team inspect the work across time zones.
Devpost's online teamwork guide says virtual hackathons use Participants tabs and Slack or Discord-style community channels for team formation. That is close to remote work: strangers, async context, short windows, and written coordination.
Know the in-person caveat
MLH's member event guidelines strongly encourage fully in-person hackathons and say Hacker Census results consistently show hackers prefer in-person events. It also says MLH has limited slots for digital and hybrid hackathons.
So do not pretend remote hackathons are automatically better. They are useful for a specific career goal: proving you can coordinate, document, submit, and demo when the team is distributed.
Use hybrid events as training data
DevNetwork AI + ML Hackathon 2025 combined online hacking from May 12 to May 29 with in-person awards on May 28 to 29. That mixed format is a good model for remote AI teams: async build work, then compressed live evaluation.
Your portfolio should mirror that. The async artifact should stand alone, and the live conversation should add judgment. If the project only makes sense when you personally walk someone through it, it is not remote-ready yet.
Completion is the hard part
Opportunity Hack's remote and hybrid era showed completion problems during the COVID years, then recovery after more flexible event design from 2023 onward. Its report also gives a broader completion-rate benchmark around 26.5 percent across the dataset.
That matters for remote jobs. The hiring signal is not just that you joined an online event. It is that you finished, submitted, documented the build, and can explain how the team handled ambiguity without sharing a room.
Build the portfolio like a handoff
Devpost's project planning guide tells builders to review requirements, brainstorm, build, deploy, test, create a demo video, and submit before the deadline. That sequence is also a strong remote portfolio checklist.
For each project, include what the model did, what data it used, what failed, what you would test next, and which part you personally built. Remote teams need clear ownership. A vague team win is weaker than a precise contribution note.
Apply with proof, not adjectives
Devpost's online teamwork guide says teams should respect time zones and different communication styles. Turn that into application evidence: links to async decisions, short technical notes, issue threads, demo scripts, and clean repo history.
For remote AI jobs, the best note is short. Link the hackathon artifact, name the role-relevant part, and explain one tradeoff. If the company builds agents, send the agent project. If it ships applied ML, send the project with evals, failure modes, and deployment notes.
< source-linked guide · dated when published >
browse hackathons