Apple Silicon performance engineering
MLX, Metal, and Accelerate. Machine-learning inference, signal pipelines, and compute-bound workloads tuned to the hardware they run on — unified memory, Neural Engine, P/E core topology, used deliberately.
Mac-native engineering, built for the Apple Silicon era.
MacOnCall is an engineering studio architecting hardened software systems on macOS and Apple hardware — from native applications to distributed compute infrastructure.
What we build
Native macOS applications
Swift, SwiftUI, and AppKit. Apps that feel like the operating system they ship on — signed, notarized, and engineered for the full Apple deployment surface from menu bar to background service.
Hardened security architecture
Secure Enclave attestation, Hardened Runtime, code signing, and SIP-aware deployment. Threat models that survive contact with hostile inputs — and the cryptographic primitives to prove it.
Distributed systems on Apple hardware
Coordination, cryptographic verification, and protocol design for compute networks and peer-to-peer systems built on fleets of Apple Silicon nodes. Production-grade, not proof-of-concept.
How we work
Security-first
Cryptographic verification beats trust. We threat-model before we code, and we document the residual risks rather than hand-waving them away.
Hardware-aware
Code shaped by the silicon underneath it. Unified memory, the Neural Engine, performance/efficiency core topology — used, not abstracted around.
Tested before shipped
Every change covered by tests, every protocol-level contract by formal verification where feasible. We don't ship code without the receipts that show it works.