C/C++ systems engineering
Production C/C++ for Linux and native applications: architecture, concurrency, memory behavior, APIs, protocol engines, modernization and performance work.
C++20/23 · Linux · multithreading · profiling · native UIProject-based B2B engineering · North Carolina
Engineering for the part that cannot be generic.
SPICA takes ownership of difficult software and integration problems where performance, hardware, networking, real-time behavior or physical signals matter.
External engineering, without the staffing pitch
A technical partner for products that touch the real system.
We are not positioning SPICA as staff augmentation. The useful engagement begins with a concrete engineering constraint: a module that must be built, a system that must become faster, hardware that must be integrated, or an architecture that must survive production.
SPICA works at the boundary between software and the thing software controls, transports, measures or transforms.
Engineering capabilities
Low-level enough to matter.
Broad enough to integrate.
Embedded + real-time acquisition
Software close to the hardware: synchronized acquisition, device integration, deterministic processing, telemetry, control paths and test tooling.
MCU/DAQ · synchronized sampling · telemetry · hardware integrationNetworking + distributed systems
Low-level transport and resilient infrastructure for systems where latency, throughput, observability and failure behavior are part of the product.
TCP/UDP · QUIC · HTTP/3 · SRT · WebSocket · multicastVideo + real-time media
Media processing from ingest through delivery: transcoding, packaging, transport, DRM integration, live monitoring and operator tooling.
FFmpeg · HLS/DASH · CMAF · NVENC/CUDA · SCTE-35 · SRTPDSP + industrial diagnostics
Signal-processing pipelines that preserve physical meaning before AI: FFT, Park transform, harmonics, MCSA, event evidence and condition intelligence.
FFT · dq0/Park · MCSA · harmonics · time-series evidenceGPU + edge infrastructure
Move compute to the right place: GPU acceleration, edge inference, bare-metal services, datacenter integration and remote operational visibility.
NVIDIA · CUDA/NVENC · edge compute · APIs · datacenterHow we engage
Define the technical boundary. Then own it.
Project scope can be narrow or end-to-end, but responsibility should be clear. We favor measurable deliverables, direct access to engineering and a handoff your internal team can operate.
Performance rescue
Profile a slow, unstable or resource-heavy system, isolate the constraint and implement the fix without rewriting what already works.
Module ownership
Take responsibility for a bounded subsystem—protocol engine, media worker, acquisition service, diagnostic pipeline or native application—from design through handoff.
Legacy modernization
Move critical C/C++ code forward while preserving behavior: architecture cleanup, safer ownership, current toolchains, tests and measurable performance.
Hardware/software integration
Connect sensors, boards, devices, network transport and software into one testable path with clear interfaces and operational evidence.
Architecture + prototype
Turn a difficult systems problem into a working reference implementation before a larger internal team commits to production scale.
Independent technical review
Review architecture, performance assumptions, failure modes or implementation choices and return concrete engineering actions rather than a generic report.
Documented engineering evidence
Our products show how we build.
These are not generic portfolio tiles. Each system exposes implementation choices, technical boundaries and the kind of engineering SPICA can apply to a client-owned problem.
C++ · GPU · 100 GbE · HTTP/3
CDN++
Private real-time media infrastructure connecting contribution, GPU processing, packaging, origin and delivery.
Open system ↗DSP · MCSA · 7-channel acquisitionPumpSpectra™
Industrial condition intelligence built around deterministic electrical evidence, synchronized acquisition and edge diagnostics.
Open system ↗C++23 · QUIC · ICE/DTLS/SRTPSPICA SFU
A native selective-forwarding core for encrypted real-time media and flexible signaling transport.
Open system ↗C++ · macOS · GTK4 · ImGuiSPICA Commander
One operational core exposed through purpose-built native interfaces for different environments.
Open system ↗Where this work fits
Industries where software has to respect physics, timing or infrastructure.
Industrial automation & rotating equipmentSemiconductor, test & advanced manufacturingNetworking, telecom & infrastructureAerospace, transportation & hardware productsBroadcast, streaming & real-time mediaEdge computing & remote operations
Start with the actual engineering problem
Tell us what is slow, fragile, missing or difficult to integrate.
We can begin with a technical discussion around the system, constraints, interfaces and measurable outcome—before discussing a larger engagement.
Diego CardonaCEO · SPICA Networks Inc.diego@spicanetworks.com+1 (828) 773-2535LinkedIn / Diego CardonaBoone, North Carolina · United States