| name | multi-event-radar |
| description | Survey upcoming events across a city or region within a date window and rank them by ICP density — which events are worth attending or sponsoring for a given target audience. Produces a ranked event calendar with attendance estimates, audience-fit scores, sponsor presence, cost-to-attend, and ROI projections. Distinct from event-radar (which extracts attendees from one event) and event-leads (which scores attendees of a single event). Use whenever a user wants to plan which events to attend or sponsor over the next weeks, asks 'what AI events should I go to next month', 'rank Devcon side events for us', 'which SF builder events are worth sponsoring', 'plan my event calendar for Q3', or wants to triage many candidate events down to a focused list. |
Multi-Event Radar
Decide which events to attend or sponsor by surveying many at once and ranking them against a defined ICP.
Workflow
- Gather inputs: target audience (e.g. "AI agent founders", "DevRel at infra companies", "voice-AI engineers"), geography (city, multi-city, or virtual-ok), date window (default next 30 days), purpose (attend, speak, sponsor, scout-talent), budget per event if sponsoring, max events selectable (default top 5).
- Build the candidate set by surveying:
- Luma / Lu.ma city pages and the discover feed
- Eventbrite tech listings for the region
- Partiful and other invite-style platforms for known builder events
- Specific known venues (Frontier Tower, Shack15, Solaris, AGI House for SF; equivalents per city)
- Conference side-event aggregators when the date window overlaps a major conference (Devcon, NeurIPS, KubeCon, AI Engineer)
- For each candidate event, capture: date, venue, host org, declared theme, capacity, registration URL, free/paid status, sponsor list if visible, expected speaker tier.
- Score each event 0–1 against the ICP across these dimensions:
- Audience fit (0.40) — declared theme + venue history + host org's typical attendee profile
- Density (0.25) — expected ICP attendees / total attendees; small high-density events beat large low-density ones
- Access (0.15) — can the user actually meet attendees, or is it stadium-format
- Reputation lift (0.10) — speaking or sponsoring at this event signals what
- Cost efficiency (0.10) — for sponsors: $/expected-ICP-touch; for attendees: hours-cost + ticket price
- Apply hard filters: events the user has already attended this quarter, conflicting dates with higher-scored events, events failing any explicit user constraints (no-paid, no-virtual, etc.).
- Produce the calendar with the top N events plus an explicit "skip list" of events that look attractive but failed scoring.
Output Format
Produce event-calendar-<purpose>-<date-window>.md:
# Multi-Event Radar — <Purpose> · <Date Window>
**ICP:** <one-line target audience>
**Geography:** <city/region>
**Surveyed:** <n> events · **Selected:** <top N>
## Recommended (in priority order)
| Rank | Date | Event | Venue | Score | Why |
|---:|---|---|---|---:|---|
## Skip List (looked attractive, didn't make the cut)
| Event | Score | Why skipped |
|---|---:|---|
## Conflict Log
<dates where two top-scored events overlap; who you'd choose>
## Budget Roll-up (if sponsoring)
| Event | Sponsor tier | Cost | Expected ICP touches | $/touch |
|---|---|---:|---:|---:|
## Next Actions
1. RSVP to <event> by <deadline>
2. Request sponsor deck from <event> by <deadline>
3. Block calendar for <date range>
Save the underlying scoring data as event-calendar-<purpose>-<date-window>.csv and share both.
Quality Bar
- Survey breadth: at least 3 different sources unless the user explicitly limits scope. Never produce a ranking from only one platform's listings.
- Every score has a one-line justification tied to a concrete signal, not a vibe.
- The skip list is required, not optional. Showing what you rejected proves the ranking is real.
- Conflicts are surfaced explicitly. Never quietly omit a conflict to make the calendar look clean.
- Sponsor recommendations require sponsor-tier and cost data. Without a published sponsor deck, output "request deck" as the next action rather than guessing tiers.
- Hot dates (week-of major conferences) get re-ranked by side-event density, not by isolated event score.
- Never recommend more events than the user can realistically attend. If the user said max 5 and you have 12 strong candidates, the answer is still 5.