why did we invest in Agroforte?

Datamint develops digital transformation solutions for complex industrial environments, combining data analytics, artificial intelligence, data engineering, and UX design to identify business opportunities and elevate the decision-making capacity of its clients. The company is more than a platform: it is a team of engineers and researchers who run the full cycle — from understanding the operational reality on the ground, to engineering the data, to designing the products that turn it into decisions. The output of this process today is consolidated in a flagship platform, A360, but the engine behind it is the combination of people, methodology, and accumulated industrial know-how.

We believe the market for deep industrial problems in Brazil — the operational complexity inside refineries, power grids, substations, and renewable plants — is structurally underserved. Brazilian SaaS and AI investment has concentrated on horizontal software, fintech, and consumer-adjacent verticals, leaving heavy industry to legacy SCADA vendors, generic ERPs, and consultancies that sell hours rather than systems. Generic LLMs cannot fill this gap: they hallucinate in environments where a single wrong decision carries financial, safety, and compliance consequences.

This is not a market that can be cracked with a generic SaaS playbook. It rewards rigorous academic depth, real industrial trust, and the talent density required to translate operator intuition into machine-readable knowledge — and Datamint brings one of the rarest expressions of that profile we have seen in Brazil. Hélio Lopes, PhD in Mathematics, spent 30+ years at PUC-Rio leading the data science and computer graphics labs and has worked side-by-side with Petrobras since 1992. Tulio Ribeiro brings 20+ years of experience scaling Lanx into a leading Brazilian technology services firm. Around them sits a bench of 30 PhDs and Masters from PUC-Rio — a level of talent density that is uncommon in Brazil and essential to durable infrastructure rather than just features.

This team's output is A360, an operational layer above SCADA, ERPs, and OT systems — a System of Action for plant managers, supervisors, and engineers. A360 unifies sensor data, manuals, regulation, and ERP context into a single explicit decision layer, powered by AI agents that operate on top of a domain ontology and continuously learn from every operational cycle. The proprietary MINT methodology and a multi-modal database (graph, time-series, geospatial, and document data) together deliver an architecture purpose-built for environments where hallucinations are not acceptable.

Datamint applies AI where it creates clear and measurable impact: lower maintenance costs, higher asset uptime, and audit-ready automation of decisions previously trapped in the heads of senior operators. Early customer results — material savings on Petrobras refineries through Smart Tocha and a 5.77% efficiency gain in a solar plant pilot — validate both the ROI and the urgency of the problem.

We invested in Datamint because companies that turn operator intuition into explicit, auditable, and reusable digital infrastructure in regulated industries can capture disproportionate long-term value. Datamint is well positioned to become a foundational Operational Intelligence layer for critical infrastructure across Latin America and global heavy industry.