Problem
In Brazil, there are more than 306 a thousand open arrest warrants (BNMP/CNJ) and 218 people disappear every day, while police forces operate with a deficit of 44,8% of cash (FBSP). It's not a lack of policing, it's the gap between the volume of people approached and the speed with which any human check can cover.
Add to this the fragmentation: each state maintains its own banks for wanted and missing, isolated from each other, without a national system connecting them.
TOFIND solves this gap on two fronts: it brings vector biometric identification to the police's pocket in the field, online or offline; and it integrates and normalizes these heterogeneous state banks (MISP), generating unified data that does not exist today and that, treated, also become strategic intelligence for the public manager (Intelligence).
Solution
TOFIND solves the problem in three connected layers:
Sentinel brings vector biometric identification to the police's pocket in the field online or offline, without relying on manual checks or fixed cameras. Every result is a probabilistic indication, never an automatic decision: final confirmation is always human and expert, with an immutable audit trail (hash-chain SHA-256) at each event.
MISP attacks fragmentation at the root: it connects and normalizes, via AI, the heterogeneous databases of each state, now isolated from each other, without requiring them to give up sovereignty over their data.
Intelligence transforms this unified flow into strategic intelligence for public managers: support for resource allocation and measurement of results, something that does not exist today in a standardized way between states.
Entry is via legal tender (Law 14.133/2021 or Legal Framework for Startups), with a free pilot, on-premise/private cloud infrastructure under state control, and LGPD as a sales differential.
Business model
B2G, with an average ticket of 300 reais per user.
Market
Brazil (primary market): TAM estimated at R$ 3 billion+/year, considering 27 federal units (26 states + DF) and the Federal Government as potential contractors, in packages of ~1.000 Licenses a R$280–380/month. This is backed up by real data: the CESec (Panopticon) shows that Brazil has spent at least R$ 2,65 billions in facial recognition for public safety since 2019, money proven to be in motion, now sprayed between fragmented solutions, without an integrated national player.
Latin America: There is no segmented and reliable market data specifically for tactical facial biometry in public safety in the region; available reports deal with electronic surveillance broadly. Our reading is one of opportunity, not of a closed size: countries with similar structural challenges (fragmentation of data, deficit of police personnel) represent a natural expansion of the model, foreseen in the roadmap of 3 years after consolidation in Brazil.
Competitors
Today there is no direct competitor that replicates the complete TOFIND model (mobile, offline app with a multi-state ecosystem of data and intelligence). The Brazilian facial recognition market for public safety is fragmented into two groups:
Private companies with fixed camera infrastructure such as Griaule and L8 Security (in the system used in Rio de Janeiro), focused on continuous monitoring of large areas, not on portable identification by the police in the field.
Internal systems built by the states themselves, such as Face PI and Lupa App (Piauí), PCA (Ceará) and isolated apps from the Amazonas and Rio de Janeiro PM, without the same functionalities and functions as seninel.
Our main competitive risk, therefore, is not a rival company, but the state choosing to develop internally. It is precisely this fragmentation that TOFIND solves: national standardization, offline mobile identification, and data integration that today does not exist between states.
Competitive differentiation
The difference between TOFIND is that it solves the problem from end to end, where competitors cover only one piece:
Real mobility: while Griaule and L8 Security rely on fixed cameras, TOFIND works on a police officer's cell phone, in any street approach, even without internet, a differential that no public security player in Brazil offers today.
Legal security as a product: mandatory human confirmation in 100% of the cases, without automated decision, and auditing with hash-chain SHA-256 immutable responding to the criticisms of racial bias and lack of transparency surrounding the sector (CeSec, DPU).
National Integration (MISP): Instead of each state building its isolated solution (as Piauí, Ceará, and Amazonas have already done), TOFIND connects and normalizes these banks, with the state maintaining sovereignty over its data.
Intelligence: transforms identification into strategic data for public managers, something that no competitor offers today in the same ecosystem.
Entry barrier
The TOFIND barrier is regulatory and contractual, not just technological:
Long-term public contracts: signed by law 14.133/2021 or Legal Framework for Startups, they last 1 the 3 years, extendable to up to 10 without a new tender. Winning the first contract in a state tends to lock the door on competitors for years.
MISP integration as a cost of exchange: by normalizing a state's legacy banks, TOFIND becomes the infrastructure layer on which it operates. Changing suppliers later requires redoing this integration from scratch.
Compliance as a specialty: NIST FRVT, ISO 27001 and certifications from IN GSI/PR no. 8 raise the technical and legal barrier, few facial recognition startups are born thinking about public safety compliance.
Data sovereignty as political trust: the federative model (state maintains full control of its data) solves the main objection to centralized solutions replicating that trust requires time, not just technology.
Traction
not