One industry, done deeply
Electrical panel and switchgear manufacturing. Not a horizontal platform with a manufacturing module.
India-first
Rupee pricing, GST and HSN codes in the data model, WhatsApp as a real sales channel.
Your data is yours
Per-company isolation, role-based access, full export. Your data is never used to train models.
The problem we kept seeing
Spend time in electrical panel manufacturing and the same pattern appears in almost every workshop. The business is genuinely good at building panels. It is losing money somewhere else entirely — in the days it takes to return a quotation, in the job stalled for one component nobody knew was short, in the margin that quietly evaporated between the estimate and the invoice.
The tools available did not help. Generic ERPs were built for repetitive production and required months of configuration and a consultant to describe a business that makes something different every week. Accounting packages recorded the outcome without touching any of the decisions that produced it. So the real work stayed in spreadsheets, inboxes and WhatsApp groups, and everyone accepted the drift between them as the cost of doing business.
What we decided to build
Zaptiz is deliberately narrow. It assumes you make electrical assemblies to order, that a drawing or a photograph is where each job begins, that your bills of material are created per job, and that you file GST. Because those assumptions are baked in rather than configured, a shop can be quoting real work within days instead of months.
The second decision was to put AI where the hours actually go. Not a chatbot in the corner, but a bill of material generated from a drawing, supplier price lists read from a distributor's PDF, and eleven agents doing the patient chasing — the overdue invoice, the cold lead, the shortage against next week's orders — that a two-person office never has time for.
How we think about AI and trust
Anything with an external consequence produces a draft that a human approves. Procurement proposals, requisitions, quotations and collection emails all wait for a person. Agents that only read and report run freely, because they cannot do harm.
Every AI run is logged with what it did and what it cost, credits are metered on real consumption rather than a flat fee, and your business data is used to answer your questions only — never pooled across companies, never used to train models.
Where we are honest about limits
Zaptiz is a poor fit for high-volume repetitive production, for businesses needing shop-floor machine integration, and for companies filing taxes outside India. Direct e-invoicing IRP integration and a Tally connector are not shipped today.
We would rather lose a sale than have a customer discover a gap in month two. If something you need is missing, ask before you buy — the answer will be straight.
Frequently asked questions
Who is behind Zaptiz?
Zaptiz is built by Zaptiz Technologies, a software company in India focused on one industry: electrical panel and switchgear manufacturing. We are a small team, which is why the product is deep in one trade rather than shallow across many.
Where is my data stored?
On managed infrastructure with per-company isolation enforced at the query layer, encrypted connections, and role-based access control within each company. You own your data and can export it at any time.
Is Zaptiz used in production by real manufacturers?
Yes — the platform runs live customer workloads, and the roadmap is driven by what those businesses hit in daily use rather than by a feature list written in advance.
How do I get support?
There is an in-app support desk where an AI assistant handles tier-one questions and escalates to a human in the same thread. Paid plans add email and chat support with guided onboarding; Pro adds priority support and a dedicated onboarding specialist.