Partybot

Intelligent Party Control Unit, 2014

Case study Outright Mental (self-directed)2014
Role
Concept, engineering & industrial design
Target
Give a piece of software a body: a texting robot party host that collects photos into a live album, routes song requests to the DJ, and quietly flags trouble to the hosts
Constraints
2014 · self-directed · no app install, no QR code, no account — the only infrastructure allowed is the one every guest already carries: SMS, at a real Brooklyn phone number
Force fielded
Nick Charney Kaye — concept, full-stack engineering, and the industrial-product treatment around it
Time-to-dispatch
Built and deployed at real parties in 2014; the phone number is now honorably retired, the source is not
Result
A working SMS party host — live photo album, DJ request routing, and host alerts including there's a creeper in the kitchen, taken seriously before most social software admitted the problem existed. Full-stack MongoDB/Node/Express/Angular build with a test suite, published open source
Principal
Outright Mental (self-directed)

They’re the studio’s first experiment in giving a piece of software a body.

The interface was a postcard. Small cards scattered around a party, Partybot’s photo on one side, and on the other, instructions in the first person: Hi. I’m Partybot. Text me to learn about my functions. No app to install, no QR code, no account — just the one piece of infrastructure every guest already carried in 2014: SMS, pointed at a real Brooklyn phone number (now honorably retired).

The functions

QUICK START GUIDE · PARTYBOT
Photo album, live. Text Partybot a photo and it lands in the party's album, assembled on the fly — the evening documenting itself while it happens.
DJ requests. Song requests route to the DJ without anyone leaning over the decks to shout a title into a headphone ear.
Host alerts. Anything else — we're out of ice, the door's stuck, there's a creeper in the kitchen — routes quietly to the hosts. Partybot took the creeper problem seriously before most social software admitted it existed.

The body

The feature image is the joke played completely straight: Partybot as shipped product, unboxed — wooden crate stenciled INTELLIGENT PARTY CONTROL UNIT, a wall-mount enclosure of distinctly telecom-surplus pedigree with a speaker grille and a stenciled PARTY, a netbook running the brains, quick-start and safety-and-compliance booklets, mounting hardware, branded USB stick. Software with a body gets the full industrial-product treatment, safety warnings included, because a robot that hosts your party deserves better than being just an app.

Under the hood: a full-stack JavaScript build — MongoDB, Node.js, Express, Angular — with a proper test suite, published as open source. For 2014, a texting robot with CI-grade engineering behind it was a deeply unreasonable amount of rigor for a party trick, which is exactly the studio’s kind of unreasonable.

The manifesto values direct experience — in life, in design, and in play. Partybot was built on the observation that the best party technology is the kind nobody has to look at: heads stay up, the album builds itself, the DJ hears you, and the hosts find out about the creeper. The software had a body so the people didn’t have to have screens.

The lineage. A decade before Conversation Simulator gave software a voice and Fever Tilt put a game inside a physical arcade table, Partybot was the first time the studio asked the question it’s still asking: what happens when software shows up in the room?

Proof: source on GitHub — MongoDB, Node, Express, Angular, and the whole doc folder.

Other Interactive Projects