~/ Sepehr Mahmoudian
I build agentic systems and work across continual learning, knowledge graphs and retrieval-augmented generation (RAG), with roots in computational neuroscience, information theory, neurorobotics and 3D vision. Mostly Rust and Python.
Focus
Selected work
- engram : Engram Neural Modeling Labs is a public placeholder for computational-neuroscience, spiking-network and neural-modeling experiments; its repository says the code will be open-sourced after publication.
- NCP : NCP is a versioned, project-agnostic canonical-JSON contract connecting neural simulators or neuromorphic controllers to robots, UAVs, simulators and read-only analysis clients. It has a Rust reference, an independent TypeScript validator/client and Python/C/C++ bindings; HEAD is an unreleased, release-blocked 1.0 candidate, while v0.8.0 remains the latest immutable release.
- prisoma : Prisoma is a public Rust/Python research toolkit for auditable experiment semantics and intervention-grounded diagnosis of Vision-Language-Action policies, building provenance-complete capture-intervention-replay infrastructure. PID is a conditional diagnostic behind explicit gates; the 0.9.0 release candidate is a source/research preview, not a scientific-results or production release.
- crebain : CREBAIN (Adaptive Response & Awareness System) is a research-only Tauri prototype for Gaussian-splat scene visualization, simulated cameras, ML object detection, multimodal 3D tracking and ROS/Gazebo drone simulation, built with Tauri 2, React 19 and Rust. The 120 Hz physics and ROS capabilities remain in progress.
- melkor : Melkor is a C++17 3D Gaussian-splat toolkit for deterministic conversion, inspection, geometry-based scene completion, viewing and explicit external reconstruction pipelines, with CPU/Metal and optional CUDA backends. v2 hardening is in progress; no supported production binary release is available.
- galadriel : Galadriel is an experimental safe-Rust cross-sensor consistency monitor for multi-sensor fusion, combining NIS/CUSUM, sign-preserving correlation over producer-attested projections and optional PID diagnostics. v0.9.0 is a pre-1.0 supervisor-review source release; evidence is synthetic/component-level, not field validation.
- pid-rs : pid-rs is a safe-Rust library for shared-exclusions Partial Information Decomposition and mutual-information estimation: categorical SxPID, KSG MI and default-off experimental continuous shared-exclusions/PID surfaces. v0.9.0 is a GitHub-only source-review prerelease.
- haldir : Haldir Gate is an experimental Rust mission-authorization reference monitor for NCP: it validates signed controller intents, deployment and mission admission, and deterministic policy, then prepares Gate-owned plant commands and signed decision receipts. Its 0.9 review remains NO_GO; it is not production-ready or airworthy.
- manwe : Manwe is an alpha airspace-perception research and validation workbench spanning vision, audio, multi-camera geometry and multi-target tracking, with a Python numerical/training package and Rust/Candle inference benchmarks. It targets candidate outputs for systems such as Crebain but currently ships no drop-in Crebain adapter.
- cortexel : Cortexel is a pre-1.0 TypeScript library and CLI that validates strict declarative JSON requests for neural-simulation figures and renders deterministic, accessible SVG artifacts plus exact-value tables, with fail-closed provenance. v0.9.0 has no stable published package or DOI.
- relief-atlas : relief-atlas is a Python generation pipeline and 10,079-item corpus of AI-generated GLB meshes for disaster relief, humanitarian aid and civil protection; licensing is recorded per asset.
- cobot-atlas : cobot-atlas is a Python generation pipeline for a public MIT dataset of 2,023 unique glTF 2.0 Binary meshes (2,024 GLB files, 33.5 GB) for robot/cobot simulation, manipulation research, VLA training and benchmarking. Hugging Face dataset · Dataset DOI · Pipeline DOI
Writing & research
- [Re] Measures for investigating the contextual modulation of information transmission : ReScience C 6(3), 2020 · reproducibility repository.
- Google Scholar : publications.
- Substack : notes & essays.
FAQ
Who is Sepehr Mahmoudian?
An AI/ML engineer based in Berlin, Germany. He builds agentic systems and works across continual learning, knowledge graphs and RAG, with a background in computational neuroscience, information theory, neurorobotics and 3D vision. GitHub member since 2014.
What does he work on?
Agentic engineering, continual learning, knowledge graphs and RAG; neural networks, neurorobotics and information theory; and 3D vision and scene reconstruction, primarily in Rust and Python.
What are his main projects?
His selected work spans computational neuroscience, neural-control protocols, intervention-grounded VLA-policy diagnosis, spatial awareness and airspace perception, statistical consistency and authorization monitors, 3D Gaussian-splat tooling, deterministic scientific visualization and synthetic 3D datasets. Each project above links to its public repository and evidence.
Has he published research?
Yes, including the ReScience C replication [Re] Measures for investigating the contextual modulation of information transmission. Publications are listed on Google Scholar.
Where is he based and how can he be reached?
Berlin, Germany. By email at sepmhn@gmail.com, or via the links below.