A travel app powered by a team of seven AI agents
Wendir replaces the messy stack of TripIt + Splitwise + Google Docs + the WhatsApp group everyone hates. The work is done by seven named AI specialists who collaborate, hand off tasks, and sometimes disagree in front of you.
How it works · the journey
A trip in Wendir is a workspace shared by your crew. Each phase surfaces different agents and different tools — but the data is one continuous thread, owned by your group, never spread across seven apps.
One person creates a trip. Everyone they invite gets full access for free — no per-seat paywall. The trip becomes a real-time workspace shared across iOS, Android, and Web.
Paste any link — TikTok, Instagram, Maps, blog — and Scout extracts every place mentioned. Local immediately verifies hours, crowds, and access. Every idea lands in a wishlist with a confidence pill and a citation.
Each idea gets up/down votes from every traveler. When 80% agree, it gets a “consensus” badge. When two agents disagree on details, both positions show and your crew picks.
Drag accepted ideas onto days; the map auto-sequences by geography and opening hours. Local proposes alternate layouts (sunrise rhythm vs slow rhythm vs culture-first). You pick.
When a hotel or activity makes the cut, Booker prepares a transaction intent and waits for your explicit tap. No autonomous spending — Mastercard Verifiable Intent makes the consent cryptographic.
Flight delayed 3h? Reshuffler proposes a Day 1 fix in seconds, coordinates with Booker on rebookings, and pushes a one-tap accept to the whole crew. Need a quick veg dinner near you? Concierge answers — citing sources, even by voice.
Moneybags simplifies 12 expenses into 3 transfers (PayID / Wise rails). Memorykeeper assembles a shared highlight reel from the crew’s photos with a timeline and a map. The trip ends; the memory persists.
Meet the crew · seven specialists
Wendir isn't “an AI.” It's seven AI specialists, each with a name, a personality, a tool kit, and a defined remit. Splitting the work this way is the whole game — a single chatbot that hallucinates an opening time loses your trust forever; seven specialists with cited sources and visible disagreement track closer to how real travelers actually research.
Curious, fast, link-hungry. Paste any TikTok, Instagram, Maps list, or blog and Scout extracts every place mentioned — photos, hours, vibe — each with a source link.
Patient, fact-first, quietly opinionated. Local owns destination truth — opening hours, crowd patterns, transit, weather. Often disagrees with Scout on timing — by design.
Brisk, no-nonsense, PayID-fluent. Adds expenses, splits equal/share/custom, OCRs receipts, converts currencies, and simplifies 12 expenses into 3 transfers.
Careful, double-checks, hates surprises. The only agent that touches money — prepares booking intents, never charges without your explicit tap-through.
Warm, present, always nearby. Your on-trip co-pilot. “What’s open for veg dinner near me right now?” — answered in seconds, cited, with directions. Voice-enabled.
Calm under pressure. When a flight delays 3 hours, has a fix drafted before you finish your coffee — coordinating with Booker on rebookings, Local on revised hours.
Nostalgic curator, patient with photos. After the trip ends, assembles a shared highlight reel — timeline, map, the crew’s photos in chapters. Captioned. Set to music.
Together · how the crew collaborates
A planner that's just a single chatbot has nowhere to escalate. A team of agents has hand-offs, second opinions, and clear lines of responsibility. When two of them disagree, you see both positions and pick — and that disagreement gets logged so the team learns what your crew prefers.
Trust · why this isn't another hallucination machine
24% of AI itineraries recommend a closed business. 33% of travelers got false info from AI planners. The trust gap is the category, not the bug. Wendir's solution: every output carries the agent's identity, a citation, a confidence pill, and a one-tap “why this?” that opens the full reasoning chain.
Every AI-generated itinerary carries machine-readable aiGenerated: true metadata plus a visible “Generated by AI — verify before you go” badge before the EU's August 2026 deadline. New entrants ship compliant; incumbents have legacy debt.
Why now · the open lane
Splitwise paywalled their free tier in 2025. TripIt's user base is openly looking for a successor. Wanderlog's #1 complaint is the per-seat paywall. AI-only planners (Mindtrip, Layla) hit the trust ceiling around 2024. The category is bifurcating — book-led giants bolting AI onto their funnels vs. AI-native planners that fail at group ops, offline, and on-trip execution. Nobody owns the lane below.
| Wendir | Wanderlog / TripIt | Mindtrip / Layla | |
|---|---|---|---|
| Group-seat pricing | ✓ One payer, whole crew | ✗ Per-seat paywall | — Single user only |
| Offline-first in free tier | ✓ Yes | ✗ Paid tier only | — No |
| On-trip auto-reshuffle | ✓ Reshuffler agent | ✗ Manual | — Manual |
| Expense splitting | ✓ Moneybags agent | ✗ External tool | — External tool |
| Trust UX (cite + confidence) | ✓ Every output | ✗ No | — Inconsistent |
| Multi-agent specialization | ✓ 7 named agents | ✗ Not AI-native | — Single chatbot |
Wendir is in closed beta now. Phase 1 ships in Australia first, then US / UK / SG / NZ in 2026.