SlipstreamPower Tools for
Rental Portfolios

Framer, the recipe editor

Answers built from parts, not patches

Every reading in Slipstream is a recipe, and every recipe is assembled in Framer from the same set of parts: synced data coming in, steps that transform it (including calling LLMs if desired), and a docked result where you and your owners can see it. A new question becomes a new arrangement of proven parts, not a one-off feature to maintain.

Watch the Framer tourSee every recipe
  1. Bookingssynced data
  2. Open nightsin 1 recipe
  3. Count and sumin 8 recipes
  4. If / thenin 19 recipes
Open NightsCo-host only
Nights to offer
  • Airbnb guest checks out Tuesday
  • Next 2 nights open, usually $200 a night
SunMonTue$200Wed$200ThuFriSatcheck-out

Send the offer through your PMS the evening before check-out.

a part a part that asks a language modelPaths simplified; card values illustrative.
29 recipes
in the catalogue today, each one open on the canvas
39 parts
that every one of them is built from
9 call an LLM
to read reviews or check an assumption against published evidence, each a visible step
1 part, 12 recipes
one revenue calculation, read by every recipe that needs it

Why parts scaffold and patches don’t

Most software answers a new request with a new feature: its own code, its own screen, its own upkeep. Each one works, and together they don’t add up. Framer answers with an arrangement of parts the system already has, so every recipe makes the next one cheaper to build and easier to trust.

partsTime windowIf / thenFormulayour new recipeTime windowCount and sumIf / thenprovenproven

A new recipe reuses what's already proven

You can try out your ideas by dragging and dropping parts. In many cases the rules and the verdicts they rely on are already built, and already checked and working in other recipes.

Guest LoadOwner PulseReversal WatchValue ScorecardLead-Time Drift+7Booking economicschanged once

Change it once, and it's changed everywhere

One part works out each property's monthly revenue and upcoming bookings, and 12 recipes read it. A change to that part reaches all 12, instead of being made 12 times and drifting apart.

ReviewsRead reviewsFormulalanguage model“Mesh Wi-Fi first: about $1,900 a year”you decidewhat is sent

A language model is a step, not the system

9 recipes include a step that asks a language model: reading reviews, pulling amenities from listing text, or checking an assumption against published evidence, with citations. It sits in the chain like any other part, its input and output on the canvas. The model is a setting, two can be compared side by side, and the arithmetic around it stays plain enough to read. One errant email or text from an unattended bot can do real damage, so a person is always the last step: recipes recommend, and nothing reaches a guest or an owner until you decide.

Slipstreamnew releaseyour accountMomentum Monitor0.860.86sameopen it, tune it, build on it: your call

Yours, and stable

Recipes live in your account, not in our code. Shipping a new version of Slipstream never silently changes what your numbers mean, and any recipe can be opened, tuned or built on.

QuestionA one-off featureA recipe from parts
Built fromCode written for one requestParts other recipes already use
How it worksInside the product, out of sightOn the canvas, step by step
When it reads wrongWait for a releaseTrace it to the step, and change that step
UpkeepEach one maintained on its ownA change to a part reaches every recipe using it
Where it showsWherever it was builtDocked on your dashboard, shared with owners recipe by recipe

Every recipe on the board

The whole catalogue, and what each recipe catches in a working portfolio. Read down the “built from” column and the same parts keep coming back. Those marked asked for by hosts answer something hosts asked for in public, or built by hand, before it was a recipe.

  • OCT 2026

    Fee GapsAsked for by hosts

    A stay recorded without the cleaning fee the property's other bookings on that channel carry.

    • Fee gaps
    • Count and sum
    • If / then
  • OCT 2026

    Overlapping Bookings

    One booking stored twice, its revenue counted twice, or a real double booking.

    • Shared nights
    • Count and sum
    • If / then
  • OCT 2026

    Open NightsAsked for by hosts

    A guest checks out Tuesday and Wednesday night is empty: offer it to them before they leave.

    • Open nights
    • Count and sum
    • If / then
  • SEP 2026

    Momentum vs. Pace

    Growth from last year that isn't showing up in next quarter's bookings.

    • Booking economics
    • Then vs now
    • Formula
    • If / then
  • SEP 2026

    Lead-Time Drift

    Guests booking closer to arrival than they used to, flagged when they cross the 30-day line.

    • Booking economics
    • Time window
    • Trend
    • Formula
    • If / then
    • Pick a value
  • SEP 2026

    Reversal Watch

    The last four months disagree with the twelve-month trend: a turn, caught early.

    • Booking economics
    • Then vs now
    • Time window
    • Trend
    • Formula
    • Pick a value
    • If / then
  • JUL 2026

    Momentum Monitor

    Twelve months against the twelve before: gaining or losing steam, and whether to fix the stay or the price.

    • Booking economics
    • Review pulse
    • Rank in portfolio
    • Then vs now
    • Time window
    • Trend
    • Formula
    • If / then
    • Pick a value
    • Evidence check (asks a language model)
  • JUL 2026

    Booking Pace Sentinel

    Is the next 90 days booking as well as the same dates were at this point last year?

    • Booking economics
    • Formula
    • If / then
    • Evidence check (asks a language model)
  • APR 2026

    Pricing Oracle

    Your nightly rate against comparable listings adjusted for size and amenities, in dollars a night.

    • Price vs comparables
  • OCT 2026

    Reviews OwedAsked for by hosts

    Guests who left 3 to 14 days ago without a review: the list to chase while the window is open.

    • Reviews owed
    • Count and sum
    • If / then
  • OCT 2026

    Review RateAsked for by hosts

    Which stays get reviewed, split by stay length, season and channel, against your whole account.

    • Review rate
    • Filter rows
    • Count and sum
    • If / then
  • APR 2026

    Review Signals

    What guests keep mentioning across every review, read by a language model and grouped into themes.

    • Review scan (asks a language model)
    • Prompt tuning (asks a language model)
    • First that exists
    • Group into themes (asks a language model)
  • APR 2026

    SignalWeaver

    Reads review signals across 14 categories and proposes new questions worth building a recipe for.

    • Review scan (asks a language model)
    • Hypotheses (asks a language model)
  • OCT 2026

    Upkeep Cost per Guest-Night

    Septic, water and propane bills on your statements, per guest-night: what each person who stays costs to keep the house running.

    • Cost per guest-night
  • OCT 2026

    Guest Load

    Septic, water, propane and laundry follow people, not nights: how many stayed, and whether it's more than last year.

    • Guests by month
    • ledger series
    • Then vs now
    • Formula
    • Booking economics
    • Filter rows
    • Count and sum
    • If / then
  • JUN 2026

    Upgrade Strategist

    The improvement guests' complaints point to, sized against what the property really earns.

    • Review scan (asks a language model)
    • Booking economics
    • Lookup table
    • Formula
    • Evidence check (asks a language model)
  • MAY 2026

    Maintenance Dispatch

    Maintenance risk weighted by revenue and upcoming bookings, as a work list in the order to send people.

    • Upkeep risk
    • Booking economics
    • Formula
    • Evidence check (asks a language model)
  • APR 2026

    Maintenance Prophet

    Review sentiment sliding at an older home: upkeep flagged before something fails.

    • First that exists
    • Formula
    • Review scan (asks a language model)
    • Keyword match
    • Language model (asks a language model)
    • Weighted score
    • Recent counts more
    • Review pulse
    • Then vs now
    • If / then
    • Evidence check (asks a language model)
  • SEP 2026

    Owner Pulse

    Unanswered owner questions, new charges, soft pace: early signs an owner may leave, each with the action that answers it.

    • Owner signals
    • Booking economics
    • Upkeep risk
    • Formula
    • If / then
  • SEP 2026

    Owner Market Brief

    The numbers owner and co-host both read: price against comparables, the next 90 days booked, and pace.

    • Owner signals
    • Booking economics
    • Formula
  • MAY 2026

    Owner-Churn Risk

    Sustained underpricing and falling sentiment together: owners at risk of pulling the listing.

    • Price vs comparables
    • Upkeep risk
    • Repeat intent
    • Owner context
    • First that exists
    • Formula
    • If / then
  • MAY 2026

    Value Scorecard

    The value you deliver each owner, in dollars above market, revenue and upkeep, to stand behind your fee.

    • Price vs comparables
    • Booking economics
    • Upkeep risk
    • Host responsiveness
    • Formula
    • If / then
  • SEP 2026

    PMS ↔ Statement ReconciliationAsked for by hosts

    Bookings on the statement but not in the PMS, or the other way round, and rent the two disagree on.

    • Bookings as a ledger
    • Match two ledgers
    • Count and sum
    • If / then
  • SEP 2026

    Statement Recurring-Charge Check

    A kind of charge that is new on this statement, or one that has stopped.

    • What changed
    • Filter rows
    • Count and sum
    • If / then
  • SEP 2026

    Statement Fee-Base Derivation

    What the management fee is really a percentage of: rent, rent plus cleaning, or rent after the channel's cut.

    • Find the fee base
    • If / then
  • AUG 2026

    Website Listing Audit

    Is the home on your own website, and do its bedrooms, baths, guests and location still match the PMS?

    • Is it there?
    • Do sources agree?
    • Lookup table
    • Read a number
    • Formula
    • If / then
    • Compare text
  • APR 2026

    Amenity Extraction

    Reads listing text for amenities the PMS never recorded, with two models checking each other.

    • Keyword match
    • Two models compared (asks a language model)
    • First that exists
  • MAY 2026

    Attention Queue

    Every property's open issues, put on one scale and ranked into a single queue for the portfolio.

    • Upkeep risk
    • Price vs comparables
    • Marketing gaps
    • Formula
    • Evidence check (asks a language model)
  • APR 2026

    Portfolio Orchestrator

    Where a home ranks, the gap to your top earners, and the amenities they have that it lacks.

    • Portfolio context
    • Booking economics
    • Formula
    • Evidence check (asks a language model)

Open any of them

Level 1

See it

Every figure traces back through the graph to the raw field it came from. Follow it node by node until you hit the source record.

Level 2

Tune it

The logic is yours to change. Thresholds, weights, wording, which sources win a conflict — edited by the operator, without code.

Level 3

Build it

Assemble entirely new tools from your own raw data. Not chosen from our menu, not requested from our roadmap.

Start with recipes that work. Make them yours.

Try the demoSee plans and pricing