Self-chilling ice cream · Patent pending · Phase 0
No freezer. No fridge. No cold chain. Ever.
−6°C · 90 seconds · 90ml · room temperature storage
Build updates · early prototype access · no spam
The Target
Room temperature → soft-serve in 90 seconds
Soft-serve territory. Achieved using chemistry sealed inside the packaging wall.
Sealed endothermic reaction. Squeeze to activate. Zero chemical contact with food. Room temperature storage — no cold chain, no refrigeration, no equipment.
See how it works →Three Questions. Direct Answers.
Activation Sequence
The Packaging
Every millimetre is deliberate.
Confirmed Specifications
Vanilla · Chocolate · Strawberry at launch.
Three SKUs. One pouch architecture.
Phase 2+ expansion planned.
105 × 182mm · 90ml UHT base · ~296g · Patent pending
The Product
105 × 182mm stand-up pouch. Five distinct zones. One architecture. Three SKUs at launch. Everything needed to go from room temperature to −6°C ice cream in 90 seconds — no infrastructure required.
01 — Architecture
Select a zone to explore
02 — SKUs
SKU 01 · Vanilla
Vanilla was chosen first because it eliminates noise. Clean fat-to-sugar ratio of 1:2.75, predictable freeze behavior, no competing compounds to confuse the chemistry read. When you need to know whether the mechanism works, you start with the flavor that reduces every variable except temperature. It also happens to be the most ordered soft-serve in the world — which matters for Phase 2 retail placement.
Key Ingredients
Texture Target
SKU 02 · Chocolate
Chocolate added complexity the vanilla baseline didn't have. Higher fat (10g vs 8g), higher protein (6g vs 4g), and cocoa solids that function as a secondary emulsifier — changing how the fat system behaves under cold. The fat-to-sugar ratio shifts to 1:2.8. Cocoa butter is fully liquid at room temperature storage and partially re-solidifies at −6°C. That's not a problem to solve — it's the mouthfeel.
Key Ingredients
Texture Target
SKU 03 · Strawberry
Strawberry introduced pH. Post-formulation, the base sits at approximately 4.0–4.2 — compared to ~6.5 for vanilla and chocolate. Fruit acids hydrolyse polysaccharide chains at UHT processing temperature. The stabiliser system couldn't be replicated from the other SKUs — it required recalibration from the ground up. The formulation accounts for this explicitly, not as a workaround but as designed behavior.
Key Ingredients
Texture Target
03 — The Difference
Why it exists
This is the problem, the mechanism, and the thinking behind the work. Not a pitch. Just an honest account of what this is and why it matters.
01 — The Problem
Every cold treat you've ever had was made possible by infrastructure. A factory freezer. A refrigerated truck. A freezer aisle. A plug in the wall at home. The cold chain is a continuous, unbroken line from the moment ice cream is made to the moment you eat it. If that line breaks anywhere, the product is gone.
It works well in cities. In suburbs. Anywhere the grid reaches.
The places you most want something cold — the summit, the trail, the beach an hour from the nearest town — are exactly the places the cold chain can't go. That gap has existed for the entire history of ice cream. Nobody solved it because nowhere that mattered was without infrastructure.
Until you tried to get somewhere worth being.
The gear that gets people to those places has improved every decade. The boots, the packs, the navigation, the clothing — all of it engineered to perform without infrastructure. The food hasn't kept up.
The shift
03 — Why It Exists
The founder spends a lot of time outside. Beaches, trails, mountains, camping, long runs that end somewhere worth being.
The problem presented itself the same way it probably presents itself to most people: not as a research question, but as a moment. You push through something hard. You get somewhere. The thing you want — something cold, something that actually feels like a reward — doesn't exist in that context. Ice cream requires a freezer. The freezer requires the grid. The grid isn't on the trail.
Most people register that as a minor inconvenience and move on. The gap registered as a solvable problem.
The gear that gets people outdoors has improved every decade. The boots, the packs, the navigation, the clothing — all of it has been engineered to perform in the real world. The food hasn't kept up. That felt like an oversight worth correcting.
So the work started.
03 — The Giving Commitment
One percent of revenue — not profit, revenue — goes to an outdoor conservation partner. The logic is direct: the places Shake-Ups gets eaten are the places this money goes toward protecting. Not a generic donation. A specific, narratable loop between the product and the cause.
The framework is 1% for the Planet. Third-party accountability, a community of brands that have made the same commitment, and a name the target audience already knows from brands they already trust.
The partner hasn't been formally selected. Leave No Trace is the leading candidate — the connection between eating something on a trail and funding the organisation that protects that trail is too clean to ignore. This will be formalised before launch.
This is a stated intention, not a locked commitment. Phase 0 is honest about what it is.
04 — Stay close
You'll hear when the bench test results come in. When the first real prototype gets built. When there's something worth reporting — not before.
Early list members get first access to prototype testing when Phase 1 units exist.
Build updates · prototype access · no spam · unsubscribe any time
Questions
Hover any question to read the answer. To contact us directly, email build@getshakeups.com
Shake-Ups is a shelf-stable ice cream in a flexible pouch that chills itself. You squeeze the grip zones, shake for 90 seconds, tear open the top, and eat. No freezer at any point — not at the store, not at home, not on the trail.
The outer wall of the pouch contains sealed chambers of endothermic chemistry. When you squeeze, an inner membrane breaks and the compound dissolves — absorbing heat and dropping the chamber temperature well below freezing. That cold transfers through a barrier wall into the food sleeve, chilling it to −6°C in roughly 90 seconds.
No. Zero food contact is a hard design constraint — not a goal, a requirement. The chemistry is sealed inside the outer wall. The cold crosses the barrier. The chemistry never does.
Real ice cream. The base is whole milk, cream, sugar, and flavour — no unusual additives to support the mechanism. The shaking action creates soft-serve texture by distributing small ice crystals throughout the sleeve. It tastes like something you earned.
Three SKUs at launch: Vanilla, Chocolate, and Strawberry. Each has its own colour system, botanical illustration, and flavour-specific accent. The architecture is the same across all three.
The target cold window is 12+ minutes below −3°C after activation. Phase 0 testing is validating this against the go/no-go criteria. Real-world performance will be confirmed before any claim goes on the packaging.
No. Activation is irreversible. Once the membrane breaks, the reaction starts and can't be stopped or reset. Each pouch is a single use.
Yes — that's the point. Phase 0 includes a stress test at 30°C ambient (beach/summer real-world conditions). The chemistry needs to overcome a higher starting temperature, so hot ambient is explicitly part of the testing protocol, not an afterthought.
No launch date yet. Phase 0 chemistry validation is underway. If the test results clear the go/no-go criteria, Phase 1 mechanical development begins. The waitlist gets updates when there's something real to say — not before.
The model is ambient shelf — no freezer case required. Target channels include outdoor specialty retail (REI, trail-focused stores), ambient vending, events and festivals, DTC online, and travel. The cold chain elimination is what makes all of these viable simultaneously.
Target is 9–12 months at room temperature. The base is UHT-processed and aseptically sealed — the same technology used in shelf-stable milk. No refrigeration at any point in the supply chain.
Honest answer: the multi-layer flexible film pouch is not currently recyclable in most markets. It's required for the mechanism to work. We're acknowledging that directly rather than obscuring it. Mono-material alternatives are on the roadmap as volume and material science allow.
The spent chemistry — ammonium nitrate solution — is non-toxic and water-soluble. Safe to dispose of in a bin or, in a pinch, wash out with water. Disposal instructions will be printed on the pack. Leave No Trace principles apply.
Target retail price is $4.99–$6.99. COGS target is under $3.50 fully loaded. These are working assumptions — Phase 1 production costing will confirm whether the economics hold at the volume required to hit that price point.
A provisional patent application has been filed covering the full system architecture — endothermic chemistry, barrier wall geometry, food sleeve dimensions, activation sequence, and layer stack. Filed March 2, 2026 under micro entity status. Application #63/994,571. Interested in reviewing the filing? Request a copy or reach out to build@getshakeups.com.
How It Works
No compressor. No electricity. No cold chain. Every degree of cooling comes from a sealed endothermic reaction inside the pouch wall — triggered by one squeeze, sustained by geometry.
01 — The Layer Stack
The pouch isn't packaging — it's the product. Every layer has a specific thermal or structural role. Cross-section through the 105mm width — bilateral symmetry, food sleeve at centre, PE foam insulation on both faces.
Pouch wall cross-section · tap a layer
02 — Inside The Pouch
Interactive cutaway of the bilateral v9 architecture. Drag to orbit. Toggle activation to see the membrane rupture, solution formation, and cold transfer into the food sleeve.
03 — The Trigger
Squeezing the 12mm bilateral grip zones raises internal pressure. A membrane sealing dry salt from water ruptures. The endothermic reaction fires immediately, and the cooling chamber begins dropping.
Before — Sealed
After — Activated
04 — The Texture Science
Industrial soft-serve machines use a scraped-surface heat exchanger — a rotating drum that sweeps ice crystals off a cold wall and suspends them through the bulk. Result: a dense network of crystals <50μm. That's the texture.
The food sleeve is flat and thin — 5 to 8mm. Every shake drags cream across the cold inner wall. Each pass scrapes off micro-crystals and distributes them. 20–25 shakes over 90 seconds. Same result as an industrial machine.
Shaking introduces 15–20% air overrun simultaneously. Soft-serve isn't just cold cream — it's aerated cold cream. The geometry does both at once.
05 — Temperature Timeline
Temperature at the food sleeve from activation through consumption. Chemistry checks out on paper — Phase 0 bench testing will confirm exact curve.
06 — Zero Food Contact
The food sleeve is sealed independently — filled, heat-sealed, then encased inside the outer pouch assembly. Chemistry sits on either side. The isolation barrier is a physical layer between them, not a distance assumption.
Cold transfers through conduction across the barrier wall. No opening. No shared chamber. The PE inner surface of the food sleeve is the only thing the ice cream ever contacts.
Story
Brand story, origin, and the people behind the build.
The Build
From bench chemistry to retail shelf. Each phase has a gate. Nothing advances until it passes. Phase 0 is active now.
Select a phase to explore
"The chemistry either works or it doesn't."
That sealed endothermic chemistry can cool a food proxy from room temperature to well below freezing, hold that temperature for 10+ minutes, and produce no meaningful odour. The tiered assessment determines how aggressively to optimise the other four levers (sleeve geometry, base formulation, thermal pathway, activation mechanics) in Phase 1.
Food-grade, non-oxidizer path. Lower cooling power than NH₄NO₃ but radically simpler commercially. One definitive test to close the door or open further exploration.
Ammonium nitrate alone. Highest solubility, highest cooling energy per volume. Four loading variations including the exact v7 pouch ratio. This is where the strategic data lives.
Two-stage cascade. Only triggered if Arm B lands in the yellow zone (−4 to −8°C). Must beat single-stage by ≥5°C to justify the added complexity.
No capital committed to mechanical engineering until this gate passes. That's by design — Phase 0 exists to fail cheaply if the mechanism is wrong.