Engineers & Technologists
We're building a real-time platform that transforms collective biometric data into generative sound — with sub-100ms latency, adaptive learning, and no existing blueprint to follow. If that excites you, keep reading.
The technical challenge
Emphonic ingests real-time biometric signals from groups of 2–30+ participants, aggregates them into collective states, and renders those states as generative spatial audio — all within a latency window that preserves the felt sense of cause and effect.
The system learns across sessions, adapts to group composition, and operates through an event-driven architecture with four processing tiers — from emergency response at <20ms to background learning at <10s.
Ingesting, validating, and stabilizing data from EEG headbands, cardiovascular monitors, and respiration sensors in real time. Signal quality scoring, fault tolerance, and graceful degradation when sensors drift or disconnect.
Synthesizing individual physiological streams into group-level representations — mapping valence, arousal, and coherence across a dynamic participant set without privileging any single signal.
Real-time spatial audio generation where collective states drive tone, density, rhythm, and spatial distribution. The sound must feel responsive without becoming reactive — a field, not a dashboard.
Session-over-session pattern recognition that refines system behavior without losing sensitivity. The system remembers, but never repeats — evolving its responses while staying attuned to the present.
Current technical foundations
The conceptual framework and system architecture are established. We're now building the first functional prototypes across the sensing, aggregation, and translation tiers.
Empatica EmbracePlus, Muse EEG. Modular sensor interface with quality scoring and privacy controls.
Python and C++ pipelines for low-latency biometric analysis. Event-driven architecture with four temporal tiers.
Max/MSP with Ableton Live. Spatial audio rendering and generative synthesis driven by collective state parameters.
Emotional processing across valence, arousal, and coherence dimensions. Pattern recognition and contextual adaptation.
Seven modular components with standardized interfaces. Parallel development, scalable deployment, non-disruptive maintenance.
Who we're looking for
We don't expect anyone to match every dimension of this system. We're looking for engineers who go deep in at least one of these areas and are genuinely curious about the others.
Event-driven architectures, real-time processing pipelines, latency-sensitive systems. Experience with complex system orchestration.
Real-time audio synthesis, spatial audio rendering, Max/MSP or similar environments. Comfort working at the boundary of engineering and sound design.
EEG, HRV, respiration data. Sensor integration, signal quality assessment, noise handling. Understanding of what the data means physiologically.
Pattern recognition, reinforcement learning, emotional intelligence modeling. Interest in systems that learn without being trained in the conventional sense.
Platform tooling, monitoring, deployment systems. Making complex systems operable and maintainable across diverse installation contexts.
How we work
We're in an early development phase — building core system behaviors through small-scale experiments rather than scaling toward production. This means you'll have direct influence over architectural decisions that will shape the platform for years.
We work in cross-functional sprints, maintain rigorous documentation, and value the kind of engineering that stays curious about why a system should exist, not just how to build it.
Collaboration models are flexible — from ongoing contributors to focused prototype sprints on specific modules. What matters is genuine engagement with the work.
Tell us what you work on and what interests you about this system. We'll find the right conversation from there.
Get in touch