Building intelligent systems. Advancing computational frontiers.
Florianópolis, Brazil
WAR Enterprise is a Brazilian technology company focused on artificial intelligence, advanced computing, and research-driven software engineering.
We develop experimental AI models, developer infrastructure, and computational systems, connecting research with practical applications.
Our work emphasizes efficient architectures, rigorous evaluation, reproducibility, and long-term technological development.
| Initiative | Research Focus | Status |
|---|---|---|
| WARMIND | Language models and AI systems | Public experimental release |
| IPORÉ | Advanced model development and training research | Research & development |
| Trace2MCP | AI agent traces and reusable workflows | Public alpha |
| Quantum Channel Lab | Quantum simulation and computational research | Research & development |
WARMIND is an AI research initiative exploring language model development, architecture, and computational efficiency.
WARMIND-200M-V2 is a publicly available experimental model with approximately 200 million parameters.
Trace2MCP explores the transformation of successful AI agent execution traces into reusable, structured workflows.
The project aims to improve workflow reproducibility and support developer experimentation with agent infrastructure.
IPORÉ is an ongoing research initiative exploring model development, training methodologies, and the engineering of more capable intelligent systems.
Technical specifications and evaluation results will be disclosed as they become ready for public release.
Status: Active research and development.
Quantum Channel Lab investigates quantum computing concepts through simulation, software engineering, and computational experimentation.
Areas of interest include quantum algorithms, quantum information, simulation tools, and connections between quantum and classical computing.
Focus: Research software and computational simulation.
Efficiency — Research architectures and methods that use computational resources effectively.
Scientific rigor — Evaluate experimental capabilities using appropriate tests and documented methodologies.
Reproducibility — Encourage verifiable results, version control, and clear technical reporting.
Practical utility — Develop technologies with meaningful applications beyond experimental demonstrations.
Responsible development — Clearly distinguish research prototypes from production-ready systems.
WAR Enterprise participates in NVIDIA Inception, an ecosystem program supporting technology startups.
Program participation does not constitute NVIDIA validation or endorsement of individual models, research findings, or products.
We welcome conversations with researchers, engineers, technology organizations, and potential research partners interested in:
Website: https://warenterprise.com
GitHub: https://github.com/WarEnterprise
Hugging Face: https://huggingface.co/warenterprise
WAR ENTERPRISE
Independent Research. Engineering. Innovation.
Brazil · 2026