MindWalk Showcase ReefIQ™ drug discovery AI from AMD Advancing AI

AUSTIN, Texas – MindWalk Holdings Corp. (“MindWalk”) (NASDAQ: HYFT), a Bio-Native AI company, today presented the first public demonstration of ReefIQ™, its biological layer for AI drug discovery, running on AMD Instinct™ at AMD’s Advancing AI 2026 in San Francisco. MindWalk is featured in the AMD Instinct™ Demo Showcase, featuring one of four custom demo vignettes, with accompanying video chat distributed through AMD channels.

“In the life sciences of AI, the long-term benefit is not a group licensing model but a biological context to think about”

This demonstration is the first time that MindWalk has publicly shown ReefIQ in action. It walks the way ReefIQ reconnects disaggregated discovery data into the underlying biological complexity, in question, and enables AI models and agents to think in that context, running on AMD Instinct open computing. ReefIQ, launched by MindWalk in June 2026, sits between client discovery data and AI thinking workflows.

A product film introducing ReefIQ is available at: https://youtu.be/S5kVmpc-ik4What is ReefIQ Biology is connected, but drug discovery data is often fragmented across files, formats, systems, teams, and workflows. ReefIQ rewires those presentations before AI acts on them. Sequences, structures, experimental results, omics, literature, evidence, origins, and program history are governed, a questionable interconnected biological context for AI-enabled discovery.

HYFT® Technology is MindWalk’s core biological pattern technology, the foundation of biological representation with transparent relationships across sequence, structure, function, method, method, evidence, and documentation. ReefIQ™ is a client biology context layer built on HYFT. LensAI™ is a conceptual and operational layer. HYFT enables biological representation. ReefIQ organizes and manages context. LensAI uses an intuitive workflow.

Drug discovery organizations generate important context throughout the process: positive results, negative results, test conditions, design decisions, machine assumptions, literature links, and reasons for candidate non-development. ReefIQ stores that context as managed, non-query information. As data is added, the network is developed and its relationships and annotations are refined. Future plans benefit from evidence that is constantly being combined with information.

HYFT thinking is based on evolutionary constraints, which preserve biological patterns because their function is non-negotiable, rather than mathematical. In controlled drug discovery, where a false positive is a concern for patient safety and program costs, that support is what makes reasoning, including independent reasoning and practice, credible enough to be acted upon.

Bio-Native inference is data-driven and memory-aware: enrichment, search for similarities in hundreds of millions of patterns, and reasoning in a connected biological context. AMD Instinct’s memory-rich, open-stack, portable profile maps for that task. The architecture supports flexible integration across data management, model access, and computing deployment, so that the customer’s connected context stays on the customer’s premises under existing management while the concept runs on open computing.

MindWalk’s Business Development team will use Advancing AI 2026 to engage life science customers, infrastructure partners, and agent-AI developers exploring where to host and dominate the context of controlled biology. LensAI is already in contracted, recurring programs with life science customers today. Interested parties may request private information through Investor Relations. For more information about the technology, visit https://www.mindwalkai.com/reefiq.

MindWalk Holdings Corp. (NASDAQ: HYFT) is a Bio-Native AI company that builds the BioIntelligence infrastructure needed by life sciences AI and agent AI. Its proprietary HYFT® Technology powers ReefIQ™, a life science context layer, organized around HYFT pattern objects that recognize activity in sequence and structural biology and refined over 20 years of refinement of a continuous representation of the biological evolution of 660 million biological and biological patterns including 25 billion across data patterns and a system that integrates all data measurements. run to the platform. The LensAI platform runs on this infrastructure to enable targeted acquisition, candidate due diligence, and portfolio decision support, as well as host AI workflows pharma agents run to implement. MindWalk’s architecture is designed in such a way that the HYFT representation layer, not any AI model per se, is a durable, unifying competitive asset.