investment thesis
One record, not eleven integrations.
The problem is the seams, not the tools
Every tool a small business uses is individually fine. WhatsApp Web is fine for talking. A spreadsheet is fine for a pipeline. Email is fine for chasing an invoice. The failure is at the seams: the client's data lives fragmented across all of them and is current in none, so somebody spends their week retyping one system into another — and the leads that fall through the gaps never show up as a cost, because nothing measures the message nobody answered.
The standard answer is integrations: connect the tools so the data flows. That moves the problem rather than solving it. An integration is a promise that two systems will keep agreeing about something forever, and every one of them is a thing that breaks quietly, in a direction nobody is watching, usually at the moment it matters. Ten tools with ten integrations is not one system; it is ten systems plus ten new ways to be out of sync.
The bet: one record every module shares
niiko does not integrate modules — they are built on one client record. The lead the agent qualified over WhatsApp is the client in the CRM, with the whole conversation attached. The appointment sits on the same availability the agent reads. The invoice, the proposal and the delivered work all point at the same record. Nothing has to be kept in sync, because nothing was ever separate.
The economics of that compound in one direction. Each additional module is worth more to a customer than the last, because the record it needs is already populated — and it costs us less to build, because the hard parts are shared. A competitor can copy any single module in a quarter. What they cannot copy in a quarter is the record underneath it, and the ten adjacent decisions that let a new module use it on day one.
Why the agent is the wedge
The entry point matters as much as the platform. Ours is the part of the business that hurts every day and pays for itself immediately: every inbound lead answered and qualified in seconds, over WhatsApp, at any hour. It is the module a business can evaluate in a week without changing how anyone works, and it is the one that fills the client record as a side effect of doing its job. The rest of the platform becomes obvious once the record exists.
It also puts us on the right side of a shift. Software that waits for a person to open it and type is losing ground to software that acts and then reports. The careful version of that — an agent that proposes, with a boundary in code around what it may do unattended — is what we describe in the note on autonomy. We think that boundary is the durable asset, not the model behind it: models get better and cheaper, and the question of who is accountable when software acts on its own does not change.
Why here, and why now
In Latin America WhatsApp is not a marketing channel — it is where business is conducted. A product designed around it is not localised software, it is software that matches how the market already works. And the incumbent we displace is usually not a competitor: it is a spreadsheet, a notebook and somebody's memory. That is a market where the bar is not "better features" but "worth changing at all", and where a tool that answers messages while you sleep clears that bar on its own.
The timing argument is unglamorous: for the first time, an agent good enough to hold a real qualifying conversation costs cents per lead. That is a recent fact, and it is what makes the wedge viable now and not three years ago.
What has to be true for this to work
Three things. That businesses adopt an agent that talks to their customers — which means trust, which is why the guardrails are engineered rather than promised. That the second and third modules attach to accounts that came in for the first — the compounding is the whole thesis, and it is the number we watch most carefully. And that we keep the delivery rate we have had, with a team this size, without the quality drift that usually pays for speed.
The honest risks
Key-person concentration. Vorluno is founder-led and the founder writes the code. That is why the delivery rate is what it is, and it is also the single largest risk in the company. Any serious conversation about capital is partly a conversation about turning one person into a team without losing the standard.
Platform dependency. The wedge runs on WhatsApp, which means Meta sets the rules, the pricing and the review timelines. We have designed around the parts we control — consent-first messaging, our own number path that does not require app review — but we do not control the platform.
Early concentration. We are early. Revenue concentration, cohort depth and pipeline are exactly the things that need to be looked at closely rather than summarised, which is why they live in a conversation and not on this page.
Scope. Nine modules from a small team is either the compounding thesis working or focus spread too thin, and the honest answer is that the same evidence supports both readings today. Our argument is the shared record — each module is much less work than it looks because it is not a separate product — and the way to test that argument is the changelog, which shows what each one actually cost.
What we are looking for
Capital that understands multi-tenant B2B software and a long build, and that reads a public audit with 44 findings as a signal of discipline rather than a red flag. If that is you, write to [email protected]. The current numbers — recurring revenue and its build, cohort retention, acquisition cost and payback, burn and runway — come with the first real conversation, under an NDA.