CORTEX

CORTEX (COgnitive Reasoning and Task EXecution) is an enterprise-grade AI automation platform designed to bridge the critical gap between AI investment and real-world workflow execution. Modern organizations use an average of 897 applications, yet 71% remain unintegrated. Most AI solutions depend on APIs that don't exist in legacy or highly regulated environments. CORTEX solves this by serving as a "pre-frontal intelligence layer" that runs entirely within an enterprise's secure perimeter, automating workflows while ensuring data never leaves the local machine. The platform is built as a dual-system architecture: CORTEX Studio acts as the user-facing workspace where workflows are captured via screen recording, analyzed, and structured. Using telemetry-guided sampling and vision-based change detection, it extracts keyframes, converts them into atomic UI actions, and organizes them into executable pipelines. CORTEX Brain is the LangGraph-based orchestration engine that plans tasks and delegates them to specialized agents (GUI, Infra, Code, MCP, and QA). It perceives screen elements, reasons through multi-step logic, and executes precise mouse, keyboard, terminal, and system-level commands. Built on a FastAPI backend, Electron frontend, and Azure AI services (including gpt-5.4-mini for reasoning and EvoCUA-8B for visual grounding), CORTEX requires zero API integrations and features real-time observability, allowing users to monitor step-by-step reasoning, inspect execution plans, and intervene when necessary. It is purpose-built for high-impact use cases like ERP automation, autonomous QA testing, and internal tool management across finance, healthcare, and defense. Currently MVP-ready, CORTEX supports concurrent workflow execution, organizational workflow sharing, and continuous reliability improvements through feedback-driven learning, positioning it as a scalable, model-agnostic, and privacy-first alternative to brittle RPA and cloud-dependent AI agents.