About
Making the physical world countable
Our mission is to make the physical world countable.
Every bar knows what it sold. Almost none know what actually left the shelf. We build the second record — and the gap between the two is the signal.
It starts at the counter and reaches everyone downstream: the outlets that pour, the distributors that move, the brands that make.
Pour Precise. Gandivum.
Why the bar counter
A bottle is a number. So is a ticket. They rarely agree.
A 750 ml bottle nominally yields 12.5 pegs. In a real bar it might yield fourteen, or eleven, depending on who is pouring and how busy the rail is. The difference between what left the bottle and what hit the ticket has simply been priced in as a cost of doing business — invisible to the owner, because nothing in the room produces an independent record of what was poured.
Stock counts at the end of a shift tell you something went missing. They do not tell you when, where, or how.
So we put a camera over the pour zone and let it produce its own count of every drink poured, then reconciled that count against the POS ledger. Two independent records of the same event; the divergence is the signal. It is a narrow problem with an unambiguous answer, which is exactly why it is a good place to start.
Where it
leads
Now
The operational tool
Barvis in venues, counting pours and reconciling them against the till. Every figure reviewed by the operator before it is acted on. This phase is about being right, repeatedly, in rooms that have no reason to trust us yet.
Next
The same cameras, more questions
Planned for 2027. Queue depth, occupancy by zone and footfall run on infrastructure already installed and already trusted — same edge node, same cameras, same privacy posture. The marginal cost of the second question is close to zero, which is the whole argument for going deep on the first.
Then
The intelligence layer
Pour-by-pour, SKU-by-SKU data accumulating across venues is the foundation for distributor and brand-owner intelligence — the point at which an operational tool becomes a data business. Aggregated and anonymised, always: a venue’s identifiable operational data is never sold. It only works if the first phase earned the right to it.
The name
The Gāṇḍīva (ഗാണ്ഡീവം) was Arjuna’s bow — chosen for precision under pressure rather than for force. It is the right image for a measurement company: the value is in where the arrow lands, and in being honest about the spread.
We are hiring vision engineers and looking for pilot venues