Inside AI Analytics
Built to see the whole system.
AI Analytics brings independent research and systems engineering together. We develop public-data infrastructure, privacy technology, and distributed systems — connecting the investigation, the software, and the way people use it.
From investigation to infrastructure
A collective approach to complex systems.
Research informs the engineering. Engineering extends the investigation. Source collection, entity resolution, protocol design, and usable interfaces belong in the same conversation. We work across those boundaries to carry a problem from the first record to a working system.
The work spans censorship evidence, public-record intelligence, privacy applications, publishing infrastructure, and encrypted communication. Each project has its own purpose and scope. Together, they form a body of software and research that can be inspected, used, and built upon.
Research, platforms, and systems.
Voidly
Voidly is an independent company building private internet tools, open censorship intelligence, and marketplaces for AI-agent services. Its public record traces reported blocking to measurements; its product work includes Voidpay, Veil, and VPN access.
Federal Data Hub
A keyless API and a catalog of 229 datasets from government sources, regulators/SROs, and derived indexes. The engineering work joins ingestion, normalization, entity resolution, and a usable public interface.
Nexcom
A network of local and specialist publications operated by NEXcom Digital, Inc. Its shared publishing infrastructure combines source attribution, AI-assisted production, and human oversight.
Swarm SDK
Applied cryptography for autonomous-system communication. The public case study explains hybrid key exchange, session security, mesh routing, and transport integration; the implementation is proprietary.
A public body of work
Follow the result back to the method.
The site also holds institution-level accountability datasets and the technical writing behind the projects. Source links, coverage, and reuse terms stay with the relevant record.
- Federal catalog
- 229 datasets
- Accountability collection
- 16 datasets
- Technical writing
- 526 articles
- Public code
- Voidly on GitHub
Careful where it counts.
Evidence has limits.
A source record, a model output, and a verified result carry different weight. We keep those distinctions explicit. Public accountability work focuses on institutions and documented actions, with source provenance and a route for corrections.
Useful systems are explainable.
We publish methods and engineering notes so others can examine the decisions. Public datasets state their own licenses; proprietary work is identified as such. The documentation should make clear what someone can use and what a result establishes.
Internal research systems
The tools behind the investigation.
Footprint Vault is in-house research tooling for organizing public-source evidence and resolving relationships. The public methods note explores collection, attribution, and graph-based analysis. The tooling itself is internal.
Origins & ownership
Independent direction.
Clear responsibility.
Dillon Parkes leads AI Analytics and founded Voidly, an independent company. He co-founded Nexcom with Galen Kahlenberg in 2024. Nexcom is operated separately by NEXcom Digital, Inc.
Voidly company and founder · Nexcom ownership · Company facts and biography
Site delivery and technical reference
This site is a static Next.js 16.3.2 export served on Cloudflare Pages. The public APIs and data infrastructure run separately.
Have a problem that crosses disciplines?
We work on research collaborations, software development, and infrastructure that crosses technical boundaries. Bring the problem, the constraints, and what a useful outcome would look like.