Use cases · in plain language
Because things move, and nothing remembers where they went. Wherebase is a memory of what was where, when, that any camera can fill and any AI can ask. Here is what that is for, from a lost pair of glasses to a fleet of robots. Each case is one question and one answer; open it for the problem, what you do and what you get.
The problem
Keys, glasses, the remote, the charger, the kid's shoes. You know you had it. You do not know where.
What you do
Point your phone at the room now and then, or wear the pendant. Nothing to tag, nothing to type.
What you get
Ask your assistant, out loud or in chat: “where are my keys?” You get the last place it saw them, and the place before that, which is often the one you need.
Who it is forAnyone who loses things. People who cannot see. Households with more than one pair of hands moving things around.
The problem
Insurance wants a list. Moving wants a list. You have never made the list, because making it is miserable.
What you do
Walk every room once with the camera on. Each thing it recognises is stored with the room and the time. Do it again next year and the differences are the story.
What you get
A list by room you can ask for in plain words, export, or hand to an insurer. Not photos in a folder: rows you can question.
Who it is forHomeowners, renters, anyone with a garage, a studio or a storage unit.
The problem
Every turnover is a guess: what is missing, what moved, what broke. By the time you notice, three guests have passed.
What you do
A short walk at every checkout with the phone. Thirty seconds per room. The memory keeps each walk as its own moment in time.
What you get
“What changed since the last checkout?” gives you what is new, what moved and what has not been seen since. Evidence with a timestamp, not a feeling.
Who it is forHosts, co-hosts, cleaning teams, property managers with more than one unit.
The problem
Returns are disputes waiting to happen: the child seat, the charging cable, the parcel shelf, the scratch nobody remembers.
What you do
A walk around and through the car at hand-out and at return, same phone, same thirty seconds. The memory keeps both.
What you get
Ask what was in the car at hand-out and what came back. The items are listed with the time they were last seen, so the conversation at the desk is short.
Who it is forRental counters, car-sharing operators, company fleets, anyone lending a vehicle with things in it.
The problem
Claude knows your space in one chat and forgets it in the next. ChatGPT never knew. Your own agent runs somewhere else entirely. Memory trapped inside one model dies when you switch.
What you do
Connect each assistant to Wherebase once, over MCP. They all read and write the same rows.
What you get
Tell Claude where the charger is on Monday; ask ChatGPT on Friday; your home robot acts on it on Saturday. Switch models, keep the memory.
Who it is forAnyone using more than one assistant. Developers building agents that must survive a model change.
The problem
For someone who is blind or visually impaired, “where did I put it” is not an annoyance, it is the day. A phone in the hand is not an answer; a camera on the chest that remembers is.
What you do
Wear the pendant. It sees as you move, remembers quietly, and answers out loud when you press the button and ask.
What you get
Answers in words a person can act on: “on the kitchen table, to your left”. Works offline for the questions it has heard before. A carer can see the same memory, if you want them to.
Who it is forBlind and visually impaired people, and the people around them. This is the use case the whole thing was built for.
The problem
A parent living alone, a relative after a fall. You want to know the day is normal without watching them.
What you do
A pendant, or a camera in the rooms that matter. Objects only, never faces: the walker, the pill box, the kettle.
What you get
A timeline of what moved and what did not. The pill box that has not moved by noon. The walker still in the hallway at night. The person at home can ask the same memory out loud.
Who it is forFamilies, carers, assisted-living teams. Dignity first: it is a memory of things, not surveillance of people.
The problem
Shared benches eat tools. The oscilloscope probe, the good tweezers, the one working USB-C cable. Searching is the main activity.
What you do
A phone walk at the end of the day, or a fixed camera over the bench. Everyone writes to one memory.
What you get
Ask from the terminal, in Claude Code, where the thing is. Ask what is usually on bench three on Mondays. Ask what left the room this week.
Who it is forMakers, hardware teams, research labs, repair shops, film and photo studios with kit.
The problem
A robot that forgets where it saw the pallet repeats the search every shift. Ten robots that do not share what they saw are ten separate guesses.
What you do
Each robot writes sightings with place and time into one space. Any of them, and any planner, can ask.
What you get
“What moved overnight in aisle three?” is one call. “Have I seen this before?” is a number between zero and one, which is what a safety monitor needs.
Who it is forWarehouse and inspection robotics, drone operators, AMR vendors, research groups with a fleet of two.
The problem
Meeting rooms, hot desks, equipment cupboards: things that belong to everyone belong to no one, and vanish.
What you do
A camera in the room, or the facilities team's phone on their round.
What you get
Ask where the remote, the HDMI adapter or the spare chairs were last seen. See what left the room and when.
Who it is forFacilities teams, office managers, co-working operators, schools.
The problem
Tools and materials move between vans, floors and days. The list in someone's head leaves with them at five.
What you do
A phone walk at the end of each day, per floor or per zone. Different people, same memory.
What you get
Ask where anything was last seen and what moved since yesterday. Fewer re-orders, fewer “borrowed” tools.
Who it is forContractors, site managers, surveyors, event crews building and striking venues.