LayrMap – Navigate, Design, Impact

LayrMap lets you design maps, guide crowds, track patients, plan events, chat, build & donate modular homes — all powered by LLMs for caregiving, safe routing, and real-time smart decisions.

Comment

Indoor environments are often confusing — especially for the elderly, patients with Alzheimer’s, the disabled, or even large event attendees. While outdoor navigation is well-supported, indoor wayfinding, caregiving support, and spatial design tools remain fragmented, inaccessible, or nonexistent.

LayrMap — a fully AI-powered, modular app that goes far beyond navigation. It brings together real-time caregiving tools, LLM-based decision support, modular home generation, event planning, and even design donation tracking — all in a single, coherent system.

Here’s what inspired each part:

Map Editor: I wanted anyone — not just designers — to upload or draw their own floor plans, tag toilets, exits, VIP zones, and route paths. It’s intuitive, and backed by DeepSeek to optimize layouts.

Navigation Engine : Accessibility matters. LayrMap generates dynamic indoor navigation with preferences like wheelchair access, quiet routes, or obstacle rerouting. This isn’t mock pathing — it’s live movement logic.

MedCare Mode: For Alzheimer’s caregivers, time is safety. LayrMap enables live patient tracking, safe zone definition, and SOS alerting — complete with an LLM-powered assistant that explains decisions, suggests safe zones, and helps families act fast.

Modular Home Designer: In many parts of the world, affordable housing needs practical tools. LayrMap generates custom passive-cooled layouts from plot size, budget, and needs — and lets you donate the design to an NGO with one tap.

Event Flow System: Crowds need control. Our event tools allow managers to pre-plan venue layouts (like queues, toilets, and stages), avoid congestion, and guide attendees with AI-assisted navigation and integrated chat.

Disaster Response Mode: Emergencies demand clarity. LayrMap lets responders quickly mark hazards, safe zones, blocked exits, and shelters on indoor maps — while AI reroutes escape paths, suggests safer layouts, and broadcasts live instructions to people inside, all in real-time and in multiple languages.

In-App Chat + Floating Assistant: Whether it’s a caregiver, event organizer, or lost attendee — everyone gets a context-aware AI assistant that explains, routes, solves, and responds live.

Donation Tracking: I wanted real-world impact. Every donated modular house design or map appears in the dashboard — showing users what lives they’ve helped shape.

Developer Focus: LayrMap is built with Bolt + TypeScript + Tailwind and follows a componentized architecture. Each page — from NavigationScreen.tsx to MedCarePanelScreen.tsx — talks to the llm.ts engine and our backend API in real-time.

This may not be a full enterprise rollout yet — but it’s more than just an idea. LayrMap is a working prototype with production-ready architecture, proving what’s possible when purpose meets design, even under constraints.

This self-access model allows users to bring their own API keys, LayrMap avoids licensing hurdles while instantly supporting global premium tools like Tavus and ElevenLabs. It empowers early adopters — especially creators, NGOs, and educators — to integrate top-tier AI features without vendor lock-in.

Privacy-first: Users own their API keys. Nothing is stored, logged, or shared — full autonomy.

Flexible by design: Core features run on DeepSeek R1, an open and production-grade LLM, but can instantly scale if users plug in premium tools.

Globally accessible: No region locks, no credit card required — LayrMap works for caregivers in rural India as much as for an event organizer in NYC.

Modular & plug-and-play: Users can enable tools like voice AI or personalized video only when needed — without bloating or breaking the app.

In the end, this wasn’t just about building a cool app. It was about merging AI, empathy, and real-world tools into something that feels alive — something that maps not just places, but human need.

Built With

  • assistant
  • bolt
  • built-with-languages:-typescript
  • chat
  • chat.ts)-design:-custom-map-&-home-layout-engines-data-flow:-json-based-zone-models
  • chat:
  • contextual
  • floating
  • javascript-frameworks:-react
  • llm-backed)
  • management
  • next.js-styling:-tailwind-css-llm:-deepseek-via-custom-llm.ts-module-backend:-custom-api-(api.ts
  • state
  • with
Share this project:

Updates