Applied mathematician · Data Scientist · AI Engineer
- 🧮 Core: Optimization via vector spaces, differential geometry, probabilistic ML
- 🚀 Currently building: netloom — virtualization of networks, optimization through geometric & algebraic structure
- ⚙️ Stack: Python, PyTorch, TensorFlow, PySpark, FastAPI, Docker, Bash
- ☁️ Cloud: Distributed platform development, high-scale systems
- 🔬 Now exploring: Low-level implementations (C/Rust) to benchmark & optimize math primitives beyond vendor libraries (BLAS, cuDNN, etc.)
| Project | Description |
|---|---|
| netloom | Cloud networking from raw Linux primitives — VPC, BGP, VPN, service mesh, all local |
| MoE-Experiments | Mixture of Experts — routing, sparsity, scaling behavior |
| Probabilistic-Deep-Learning | Bayesian neural nets, uncertainty quantification |
| Bayesian-Statistics | MCMC, hierarchical models, inference |
| Particle-Physics-Experiments | Detector simulations, event analysis |
| Domain | Tools & Concepts |
|---|---|
| Math / Optimization | Vector Spaces, Differential Geometry, Variational Methods, Convex/Non-Convex Optimization |
| ML / AI | PyTorch, TensorFlow, MoE Architectures, Probabilistic Models, Bayesian Inference |
| Data Engineering | PySpark, Distributed Pipelines, High-Throughput ETL |
| Systems / Infra | Linux, Docker, FastAPI, LVM, RAID, Networking (BGP, IPsec, VPC) |
| Cloud | AWS (VPC, EC2, S3), Distributed Platform Design, Containerized Services |
| Low-level (learning) | C, Rust, SIMD, BLAS Internals, Kernel-Level Benchmarking |