Skip to contentTHE ENGINEERING PORTFOLIO
Follow the idea.
Inspect the work.
An accepted NVIDIA PAIR contribution, the current Hermes integration path in Ares, persistent memory, and the supporting Rust systems. Each project explains what was built and where the evidence stops.
01/07
Merged upstream / September 23, 2026NVIDIA PAIR: cluster trust notifications
Duplicate endorsements and unsuccessful persistence must not announce a cluster trust change that did not happen.
THE WORKChanged the cluster manager to announce successfully saved endorsement changes. Added regression coverage for duplicates, failed writes, re-pins, concurrent submissions, and persisted read-back.
Evidence & current scope +
NVIDIA Personal AI Router PR #62 was merged on September 23, 2026 as afdf9977c4c15029523aa4e1f965b2f8cc0e559c. The accepted patch and its review are publicly inspectable.
An accepted open-source contribution to NVIDIA PAIR. This does not imply employment, sponsorship, or a commercial partnership with NVIDIA.
Inspect the accepted NVIDIA contribution↗02/07
Active development / downstream Hermes distributionAres agent runtime
A long-running agent needs a stable runtime, inspectable outcomes, and explicit control over its integrations.
THE WORKAn isolated Ares launcher and runtime control plane with immutable releases, atomic runtime selection, doctor/status checks, rollback, and opt-in RecursiveIntell integrations.
Evidence & current scope +
Source reviewed October 1, 2026 at Ares revision fadd47d7330445969cd99863e736c9b5dbfced54. The repository documents inherited Hermes surfaces, Ares-owned lifecycle code, gated execution, and external services separately. The September integration case remains available as a dated historical snapshot.
Ares is an independent RecursiveIntell downstream distribution, not an official Nous Research product. Source presence does not establish installed activation, every provider/platform combination, or production fitness.
Explore Ares↗03/07
Current core / R&D systemMnemes and semantic-memory
Agent memory must persist, retrieve relevant prior state, preserve provenance and time, and remain operator-owned.
THE WORKA Rust-first memory engine, MCP access layer, agent integration kits, and self-hosted Mnemes server with explicit ownership boundaries.
Evidence & current scope +
Public repositories, crates, documentation, tests, and a dedicated product/proof surface.
Public artifacts establish implementation scope, not customer adoption, universal correctness, or production fitness.
Inspect semantic-memory↗04/07
Public R&D componentsClaimLedger and agent-graph
Agent plans and claims lose value when evidence identity, promotion rules, execution history, and failure states remain implicit.
THE WORKPublic Rust components for claim/evidence history, typed graph execution, receipt-bearing tool paths, and bounded orchestration.
Evidence & current scope +
Published crates, public source, schemas, examples, and test surfaces that can be inspected directly.
Receipts report observed execution. They do not prove factual truth, authorization, security, or task success.
Inspect agent-graph-mcp↗05/07
Recovery / installer prototypeRecursive Linux recovery workbench
A workstation supporting long-running AI work needs a recoverable operating environment without silently taking storage, credential, or device authority.
THE WORKA Fedora 44 recovery-oriented workstation installer prototype packages bounded Wi-Fi recovery, targeted PCIe ASPM mitigation, TLP policy, disk reporting, health projection, and Hermes Workbench profiles while leaving destructive storage and identity choices interactive.
Evidence & current scope +
Public source includes a structurally checked hardware-target ISO candidate, a disposable BIOS/QEMU install-and-boot fixture, project tests, and a sanitized evidence receipt dated September 6, 2026.
The repository remains NO-GO for physical USB writing or laptop installation. UEFI installed-system behavior, physical-media validity, full workstation restoration, and hardware safety remain unverified.
Inspect Recursive Linux↗06/07
Measured research / quality gate openproveKV hybrid-state research
KV-cache and hybrid-state work needs explicit identity, inspectable formats, and evidence that separates storage wins from model-quality costs.
THE WORKA receipted two-tier content-addressed cache pool stores one shared cold prefix plus per-agent hot shells, with explicit codecs, manifests, and baseline contracts.
Evidence & current scope +
The current N=8 receipts report 40.50× versus an f32-raw baseline for the f32-radii profile and 76.54× for BlockLogU8-radii; the fp16-equivalent ratios are 20.25× and 38.27×.
The current cache-aligned continuation check also degrades from 7.203125 oracle PPL to 24.234375 (+236.44%). N=8 PPL neutrality, framework-cache reduction, decode speedup, and production fitness are explicitly not claimed.
Inspect proveKV and its receipts↗07/07
Experimentalturbo-quant and compression research
Resource-constrained AI needs smaller representations without hiding formats, quality budgets, or exact fallback paths.
THE WORKExperimental Rust quantization and compression primitives with explicit codecs, benchmark surfaces, and raw/exact escape hatches.
Evidence & current scope +
Published package records and public source for turbo-quant and related components.
No comparative performance or memory-savings claim is published here without a reproducible, revision-bound benchmark.
Inspect turbo-quant↗Good systems start
with a clear question.
Hiring for agent infrastructure? Working through a difficult system? Tell me what you’re trying to make work.
Start a conversation ↗