September 2026 Summaries

After retiring from a 33-year enterprise IT career that included CIO roles at Merck and Alexion, the author returned to biopharma as CEO of Tile.ai after concluding that artificial intelligence has renewed the longstanding challenge of enterprise data usability. They argue that prior technology waves, from ERP and cloud computing to data lakes and collaboration platforms, repeatedly created fragmented and poorly governed data environments despite substantial investment. As AI models become increasingly commoditized, the author contends that an enterprise’s competitive advantage will depend on making its distributed data discoverable, accessible, secure, and governed for both people and AI systems. Tile.ai, formerly TileDB, is presented as an Enterprise AI Data Substrate that prepares data for AI where it already resides, avoiding migrations while using a “Tile” primitive to embed data context, tools, and access controls directly into the infrastructure.

Tile.ai presents its Enterprise AI Data Substrate as a governed layer intended to make data distributed across storage systems, warehouses, SaaS applications, and specialized scientific formats available to AI agents without moving or duplicating it. The company argues that enterprise AI adoption is constrained less by model capabilities than by difficulties finding data across silos, accessing heterogeneous formats and runtimes, and applying identity-aware permissions and auditing. Its core concept, the Tile, is a wrapper around an existing data asset that provides contextual metadata, native query tools, controlled execution, federated access, and reproducible audit records while leaving the underlying source authoritative. Tile.ai positions the platform as distinct from catalogs, lakes, and warehouses because it makes data callable by agents through interfaces such as APIs, notebooks, filesystem mounts, and MCP, while retaining source-level controls. The platform can be deployed as a single-tenant installation in a customer’s cloud account, integrating with existing identity providers and infrastructure to support discovery, access, governance, collaboration, and traceability across data sources.

TileDB is becoming Tile.ai as its founder shifts from CEO to a hands-on product and vision role, while former life sciences CIO George Llado becomes CEO to lead commercial growth. The company was originally built around the idea that complex data types such as genomes, images, geospatial scans, time series, and machine-learning features could be managed through a unified multidimensional-array engine, providing cloud-native performance, open formats, and built-in governance for demanding scientific and regulated uses. The founder argues that enterprise AI has changed the central challenge from simply storing and querying complex data to making distributed proprietary data safely usable by AI agents without requiring migration, weakening governance, or relying on overly privileged service accounts. Tile.ai is positioned as an enterprise AI data substrate that connects to data where it already resides, turns assets into governed AI-ready “Tiles,” and enables agents to work across sources under users’ real identities and access controls. Currently in private preview with design partners in regulated industries, the transition has received backing from Amgen and Two Bear Capital.

TileDB has rebranded as Tile.ai and launched its Enterprise AI Data Substrate, a governed data layer in private preview designed to let AI agents securely discover, access, and use sensitive enterprise data where it resides without requiring migrations or weakening existing controls. The platform addresses data fragmentation across instruments, files, databases, and cloud services, particularly in regulated industries such as life sciences, while preserving lineage, reproducibility, and governance. George Llado, a former CIO at Merck and Alexion Pharmaceuticals and TileDB board member, becomes CEO, while founder Stavros Papadopoulos transitions from CEO to Chief Product Officer to focus on product and vision. Tile.ai also announced strategic financing led by existing investor Two Bear Capital, with participation from Amgen Ventures, to support product development, market expansion, and deployments with early adopters. The open-source TileDB array engine will continue operating under its existing name.

A Tile is presented as a governed, agent-ready interface over existing enterprise or scientific data sources that allows AI systems, applications, pipelines, and users to access data in place without copying it into new repositories. Each Tile combines a source connection and caller identity, contextual metadata and indexing, data-specific tools, and a runtime deployed near the data, enabling formats such as genomics files, pathology images, documents, warehouse tables, SaaS records, and custom systems to be queried through consistent interfaces. The approach is intended to address limitations of direct agent access to raw files and tables, including format-specific setup, repeated integrations, unclear permissions, limited discoverability, and incomplete auditability. Tiles use existing identity providers and source-level permissions, execute requests as the calling user, and record access, queries, lineage, and relationships, while supporting interfaces including SQL, REST, SDKs, FUSE, and MCP. By making each source a standardized callable object, the proposed model aims to let agents combine multiple governed sources in a workflow while maintaining access controls and avoiding what the company describes as shadow AI access paths.