Skip to content

User Journeys

Field Value
Status canonical product journey model
Owner PRODUCT OWNER / DIR
Last updated 2026-06-10
Applies to The professional-networking core loop that backend and mobile build PRs implement against.
Evidence product-spec.md; product-decisions.md PD-013 through PD-022; ../counsel-queue.md C26 through C28.

Purpose

This document is the canonical product model for what Partile should feel like to a traveler. It describes the professional-networking journey and the product constraints that keep matching tied to a verified shared place and time. It is a design/source-of-truth document, not evidence that the behavior is already implemented.

1. Product Model

Partile is professional serendipity gated on verified shared place plus time. A boarding pass proves that a traveler will be at airport X during window W. Everyone whose verified window overlaps theirs at the same airport forms the candidate pool.

The AI ranking engine orders that pool by professional relevance: complementarity, such as founder to investor, recruiter to candidate, or artist to curator; and similarity, such as same-industry networking. Discovery is an open directory of everyone in the traveler's window, relevance-ranked. Each row has a one-line "why you two" rationale. The traveler can star prospects into a shortlist, open a profile card, and request to connect with a note.

The recipient reads the note and sees the same AI rationale before accepting or declining. Accepted requests open a structured professional chat. If both sides agree to meet, Partile reveals the relevant gate or area and suggests a neutral Heathrow meeting spot. Once created, the connection persists as a durable professional contact; discovery is bound to the trip window, but the relationship is not.

Partile is not a dating app and not casual encounters. The product's first magic moment is a relevant professional introduction and conversation during a trip.

2. The Person

The user is a traveler holding a boarding pass, somewhere on the trip arc:

  • Night before / cab to the airport: scans the pass early, enters tomorrow's pool for that airport/window, and browses who will be around.
  • Post-security, coffee, lounge, or layover: checks who is here now or soon who is worth meeting.
  • Mid-trip re-open: checks for new arrivals or continues an existing professional conversation.

Professional motive is present throughout. Good outcomes include a founder meeting an angel investor, a recruiter finding a candidate, an artist meeting a gallery curator, or two relevant professionals deciding they have 30 minutes for coffee.

3. Timing Model: Place And Time

Presence is not simply "I am physically here now." It is "verified to be at airport X during window W." A night-before scan can place the traveler into the next day's airport/window pool. Matching only happens across people whose verified windows overlap at the same airport.

If two travelers will not be proximal in both space and time, Partile does not try to match them. That is why the boarding-pass verification gate is the entry point to the social graph.

4. Primary Journey: First Trip, Happy Path

  1. Sign in: LinkedIn OIDC starts the session.
  2. Profile bootstrap: Partile pulls the LinkedIn profile and may silently enrich it through PDL/ZoomInfo for the matching engine only. Enrichment is not shown on cards and is not redistributed. The traveler can sharpen their profile through quick intent tags or an optional AI-driven interview.
  3. Boarding-pass gate: presence, discovery, matching, candidates, and conversations stay locked until a verified trip exists. There is no skip and no manual flight-entry product fallback.
  4. Scan / import the pass: the traveler scans a paper or on-screen boarding pass, or imports an image/screenshot from Photos or Files. On-device barcode decoding produces a sanitized trip confirmation and verified presence. This establishes airport X plus window W.
  5. Unlock discovery: Partile shows a relevance-ranked directory of everyone in the traveler's window. The top of the list is effectively the best matches. Rows include the AI rationale, can be filtered, and can be starred into a shortlist.
  6. Open a card: the traveler sees a LinkedIn-grade profile and the match rationale. If later enabled, internal enrichment may contribute to ranking but is not displayed.
  7. Request to connect: the traveler sends a short professional note explaining why they want to connect.
  8. Recipient decides: the recipient reads the note and the same AI rationale the sender saw, then accepts or declines.
  9. Conversation: on accept, a thread opens with light professional structure.
  10. Propose to meet: either side can use a one-tap propose-to-meet action. On mutual accept, Partile reveals the gate/area and suggests a neutral Heathrow spot.
  11. Magic moment: the in-person coffee or introduction happens.

5. Secondary Journeys

  • No response / declined request: requests expire with the trip window. Declines are silent to the sender; the sender keeps browsing and the pool naturally ages out.
  • Awkward first message: the request note is the opener, so the thread does not start with a blank box. The recipient confirms professional intent to anchor the exchange.
  • Empty / thin airport: a quiet window is normal. Empty state should say that few travelers are in-window right now and may offer a conservative notify-me option when someone relevant arrives. Partile must not fabricate matches.
  • Safety moment: block, report, and report-and-block are available from any card or thread. Settings includes blocked-list management. Meeting suggestions are public, neutral, in-terminal spots only. Exact gate/area is revealed only after mutual accept-to-meet.
  • Meet-up handoff: after mutual accept, both sides see the shared gate/area, suggested spot, and practical time-window framing such as how many minutes appear to overlap.

6. Chat Structure, Intent Tags, And Notifications

The MVP chat structure is:

  1. Requester's note opens the thread.
  2. Recipient confirms intent: Advice, Intro, Hiring, Pitch, or Just connect.
  3. One-tap Propose-to-meet drives the gate/area and suggested-spot reveal.
  4. Plain text messages are otherwise allowed.

MVP chat does not include AI icebreakers, AI openers, or AI scheduling.

Intent capture starts with quick tags: Raising, Hiring, Job-hunting, Advising, Selling, and Open to connect. The optional AI interview can deepen these tags. Tags and profile data can feed ranking and the rationale; deferred internal enrichment may be added later only as a non-displayed matching signal.

Notifications should be conservative and relevance-thresholded. MVP notification types are:

  • Relevant person now or soon in your window.
  • New connection request.
  • Request accepted.
  • New message.

7. AI In V1

AI in v1 has two product roles:

  • Match rationale: a one-line "why you two" shown on each card/feed row and shown symmetrically to both sides of a connection request.
  • AI profile interview: an optional flow that deepens the user's own profile and intent signals.

AI icebreakers, AI openers, and AI meeting scheduling are deferred.

8. Explicitly Out Of Scope For V1

  • Dating or casual framing.
  • Manual flight entry.
  • Stealth or invisible mode.
  • Share-to-Partile extensions.
  • Direct Wallet pass-library access.
  • PDF decoding.
  • Multi-airport launch; Heathrow is the launch airport.
  • Realtime delivery, typing indicators, and read receipts beyond basic messaging.
  • Display or redistribution of PDL/ZoomInfo enrichment data.