| |
|
| |
Build agents that act on behalf of users
Most agents are demos. They call APIs. They look impressive. But the moment they need to act as a specific user in a real system, with real permissions, they fall apart.
This hackathon is about that last mile. Build an AI agent that doesn't just execute tasks - it does it as the right person, with the right access, in the right context.
Date: Saturday, July 25
Location: San Francisco (available on registration approval)
The problem
Connecting an agent to a CRM (or any app) is easy. Updating a record as a specific user, respecting their tenant, their permissions, their identity is where most demos break.
Today, you're solving the hard part. You'll build agents that act on behalf of real users: not as root, not as a service account, but as the person the action belongs to.
Schedule
9:45 AM Check-In
Grab your badge & get settled in for the day.
9:45 - 10:15 AM Welcome & Track Overviews
Kickoff, API walkthroughs, & track breakdowns.
10:15 AM - 4:30 PM Build Time
Hack, attend workshops, get mentor guidance. Pizza lunch provided.
4:45 PM Submissions Due
Final code locked. No extensions.
5:00 - 7:00 PM - Demos & Judging
Present to the room & vote for your favorites.
Judging criteria
Projects are scored equally across four dimensions:
Innovation (25%) - Is the approach novel? Does it solve the problem in a way we haven't seen before?
Technical Complexity (25%) - How well-built is the solution? Does it genuinely leverage all 3 tools?
Impact (25%) - Does this solve a real problem? Could it scale beyond the demo?
Presentation & Usability (25%) - Is the demo clear & compelling? Would someone actually want to use this?
Exciting cash & credits as prizes
Who should come
This is for you if you're a founder, developer, AI engineer, or agent builder who wants to work on the leading edge of agentic AI. We'll have API docs, live walkthroughs, & mentors on hand all day.
Whether you come solo or with a team, you'll leave with working code, new connections, & a sharper intuition for building agents that actually do things.
About the organizer
Scalekit is the connectivity layer between AI agents & enterprise apps - SaaS apps, APIs, MCP servers. Agent builders use our SDK to let agents act on behalf of users with per-user token lifecycle, scoped permissions, & tool-calling that holds up in production.
A couple of co-host spots are open for this one - demo slot & social media branding If that's your tool, reach out on LinkedIn.
Want to judge instead? We're looking for a few sharp people to help pick the winners. Reach out on LinkedIn if that's you.
|
|
|
|
|
|
|
|
