Blue Sky Tech × Palm Venture Studios

What you listen to
is who you are.

A behavioral matchmaking pilot using longitudinal Spotify data — the first real-world proof that what people do predicts compatibility better than what they say.

Read the pitch Get in touch
About Blue Sky Tech

A data registry
for the agent economy.

We're building a listening-data registry designed for the agent economy. Our core insight: AI agents need longitudinal, real-world behavioral signals to stay grounded — and music listening is one of the richest, most continuous preference streams humans generate.


Blue Sky Tech turns listening history into structured, cross-platform data that can be queried agent-to-agent (A2A). Instead of static profiles or synthetic preference guesses, agents reference an evolving behavioral record that captures taste shifts, patterns, and context over time.


The result: better agent reasoning, better personalization, and a path toward an open human data layer where people own and monetize the signals their lives produce. If it works for listening data, our underlying technology scales to any vertical where human behavior can be opted-in and captured.

Data type
Longitudinal
Years of continuous behavioral signal
Architecture
A2A
Agent-to-agent queryable data layer
Pilot duration
8 weeks
Scoped A/B test with clear success criteria
Pilot fee
$50K
Scope-dependent, agreed before launch
The thesis

Most matching systems ask
people who they are.

"We already know — because they've been telling Spotify for years."

Every skip, replay, and late-night listening session is a revealed preference — honest in a way no survey ever is. Blue Sky Tech's technology turns that continuous behavioral record into a structured compatibility signal.

This pilot is the first real-world test of a simple but powerful claim: longitudinal behavioral data produces better human outcomes than self-reported preferences. Better matches. Better dates. Proof that what you do is more predictive than what you say.

Control group
Matched via current logic

Existing algorithm — surveys, profiles, and behavioral signals already in the app. This is the baseline.

vs
Experiment group
Matched via music compatibility

Spotify listening history scored for taste compatibility — honest, longitudinal, zero self-reporting required.

The problem

Modern matchmaking
is flying blind.

Self-reported data
Unreliable

Users describe idealized versions of themselves and partners — not real ones. Aspirational ≠ accurate.

Survey-based preferences
Static

A 5-minute quiz cannot capture years of authentic behavioral signal. People change. Quizzes don't.

Match quality
Unmeasured

Most apps measure swipes and messages. Nobody asks whether the date actually went well. We will.

How the pilot works

Four steps.
Eight weeks. One answer.

01
Users authorize Spotify

Pilot participants connect their Spotify account via OAuth. We pull top artists, tracks, genres, and listening history — no personal data stored beyond what's needed for scoring. Cohort size scoped with the app team.

02
Generate compatibility scores

Multi-dimensional taste-compatibility score computed per user pair: genre graph overlap, audio feature alignment, temporal listening patterns. Music as a revealed behavioral proxy — not just "you both like the same artist."

03
Run the 8-week A/B test

Cohort split into control and experiment. Both groups use the app normally. After dates, both groups rate their experience. We collect, analyze, and compare. The control group sets its own bar.

04
Publish the case study

Results — win or lose — become a public case study on behavioral data in matchmaking. Blue Sky Tech retains methodology IP. The portfolio company gets a validated, differentiated feature if the hypothesis holds. The industry learns either way.

How we measure success

The vibe check.

Music is the world's best icebreaker. It should show up on the date itself — so we measure exactly that.

Primary metric

Self-reported date quality

After a match goes on a date, both users are prompted to rate the experience. The control group's ratings serve as the baseline — if the experiment group rates measurably higher, the hypothesis holds.

In-app post-date prompt 1–5 vibe rating Short 3-question check UX designed with app team
Success threshold

The experiment group's average date quality score is statistically higher than the control group's at pilot close. No arbitrary number — the control group sets its own bar. If their current method is working, we'll see it.

Investment

How we get paid.

Pilot fee (upfront)
$50,000

Covers Spotify API integration, scoring engine build, 8 weeks of infrastructure, and full analysis + case study write-up. Scope-dependent — finalized after cohort sizing with app team.

Success bonus (contingent)
To be set

Flat bonus triggered if experiment group's date quality score is statistically higher than control at pilot close. Defined and agreed before launch. Incentives aligned.

License to scale

Positive results give the portfolio company right of first refusal on a full API licensing deal. Blue Sky Tech retains methodology IP. They get a proven, proprietary matching layer — we get a validated product in market.

The ask

Three things
to get this moving.

01
Introductions to the portfolio company's product & engineering leads

We need to scope the integration, design the vibe-check UX, and finalize cohort size before locking pricing.

02
Palm's endorsement as the facilitating investor

Your trust accelerates the conversation significantly — especially given the portfolio company's second-chance capital context.

03
A defined timeline to close

Ready to move immediately. Spotify OAuth + scoring engine live within 2–3 weeks of agreement. 8-week pilot clock starts on first user authorization.

Contact

Let's build this.

Reach out directly to either of us. We're ready to move fast.

Nick Rains
Gregg Sgarlata