Founder & CEO, Genovo Technologies · NVIDIA Inception Program Systems Architect · ML Researcher · Published at AAAI '26
Lagos, Nigeria · afolabi@genovotech.com · genovotech.com · folabi.me
Temporal Eigenstate Networks: Linear-Complexity Sequence Modeling via Spectral Decomposition Accepted — AAAI-26 AIDD Workshop O(T) complexity vs O(T²) transformers · 3–28× speedup · 120× memory reduction Universal approximation proofs · Lyapunov stability analysis → OpenReview
Resonance Neural Networks: Frequency-Domain Information Processing Preprint — OpenReview O(n log n) FFT-based architecture · 4–6× parameter efficiency vs transformers Gradient stability proofs · Holographic memory encoding
Thermodynamic Manifold Alignment: Physics-Inspired Compositional Reasoning Preprint — OpenReview Statistical mechanics foundation · Parameterized manifold energy minimization SCAN benchmark evaluation
| Project | Language | What it is |
|---|---|---|
| Genovo Engine | Rust | AAA-tier game engine. 253K+ lines across 26 modules: PBR/deferred/ray tracing renderer, rigid body + cloth + fluid physics, behavior trees + navmesh AI, custom scripting VM, reliable UDP networking. |
| TocinLang | C++ / LLVM | Systems language targeting LLVM IR. Custom lexer, parser, semantic analyzer. Goroutine-style concurrency, NUMA-aware scheduling, V8 FFI. |
| TosinOS | C / x86 ASM | 32-bit protected-mode monolithic kernel. Preemptive scheduler (16 priority levels), custom paging/heap, VFS, IPC. Sub-20KB. |
| EPOCH VCS | C++ / Rust | Distributed VCS via Merkle DAGs. SIMD-accelerated diff (10× faster than Git), Ed25519 signatures, Zstd delta compression. |
| NavierFlow | Python / Taichi | CFD engine: Navier-Stokes + Lattice Boltzmann solvers. GPU-parallelized. Real-time multi-phase flow visualization. |
| Hate | Rust | Register-based bytecode VM. NaN-boxed values (8-byte), generational GC, async/await, pattern matching, closures. |
| Code Editor | C++17 / x86-64 ASM | SIMD-accelerated editor. AVX2/SSE4.2 search at 10GB/s, gap buffer, ImGui UI, <100ms cold start, <50MB RAM. |
| Emulator | C++20 | Cross-platform emulator framework for NES/GB/CHIP-8/x86 BIOS. Micro-op CPU execution, hardware-accelerated video. |
Genovo Technologies — Delaware C-Corp · NVIDIA Inception Program
| Product | Description |
|---|---|
| SynthOS | AI training data validation. 90%+ model collapse prediction. Multi-scale cascade validation across 15+ proxy models (1M–500M params). |
| Autonomous SRE Agent | LLM-driven infrastructure agent for incident detection, root-cause analysis, and automated remediation. |
| COBOL Migration Engine | Transpiles COBOL to idiomatic Go. AST-level transformation with semantic preservation. |
Production AI inference stack:
- 8× NVIDIA H100 80GB (GCP
a3-highgpu-8g,us-east5-a) - vLLM serving Qwen3-235B-AWQ on RAID-0 NVMe
- LangGraph-based agentic pipeline
Systems: Rust · C++ · C · x86 Assembly · LLVM IR · OS Kernels · Compiler Design
ML Research: PyTorch · CUDA · Novel Architectures · Spectral Methods · Mathematical Proofs
Inference: vLLM · H100 Cluster Ops · AWQ Quantization · Distributed Serving
Security: CEH (EC-Council) · Penetration Testing · Threat Detection · Cryptography
Infra: AWS · GCP · Docker · Kubernetes · CI/CD · Linux
ORCID: 0009-0001-2852-2885
Compilers, kernels, novel architectures — not frameworks.