Galaxity AI · Galaxity AI

AI & Machine Learning Systems Engineer

Build governed closed-model AI, neural-network and machine-learning systems for enterprise operations, infrastructure intelligence and autonomous decision support.

Role mandate

Build the capability for the era ahead.

This team member will combine subject-matter depth, sound judgment and disciplined execution to move the mission from architecture to measurable results.

Responsibilities

Own the Galaxity AI workstream for AI & Machine Learning Systems Engineer and deliver measurable technical or operating outcomes.
Build governed software, models or data pipelines for reliable enterprise and mission operations.
Define testing, monitoring, security and human-approval controls for production deployment.
Integrate algorithms with real operational data, infrastructure and decision workflows.

Required qualifications

Hands-on experience delivering production software, machine-learning systems, embedded code or digital models.
Strong skills in system design, testing, version control and technical documentation.
Ability to work across data, infrastructure, security and domain-engineering teams.

Preferred experience

Experience with autonomous agents, closed models, neural networks, real-time systems or hardware acceleration.
Background in regulated, aerospace or high-availability environments.

Scientific and technical requirements

Physics, mathematics and engineering foundations

Linear algebra, multivariable calculus, probability, Bayesian estimation, numerical optimization and time-series analysis.
Supervised, unsupervised and reinforcement learning; neural networks; uncertainty calibration; causal limits and model-risk controls.
State-space modeling, Monte Carlo methods, system identification and hardware/software-in-the-loop simulation.

Mathematics and working equations

Minimize empirical risk J(θ) = (1/N)ΣL(fθ(xi), yi) + λR(θ); report validation error and uncertainty, not training loss alone.
Bayes rule P(H|D) = P(D|H)P(H)/P(D); use calibrated posterior or confidence measures for operational decisions.
Reinforcement-learning return Gt = Σγᵏrt+k+1, with constrained policies, safe exploration limits and human authorization boundaries.
For digital twins, propagate ẋ = f(x,u,t), y = h(x,t) and quantify residuals between model predictions and telemetry.

Hardware, flight and infrastructure components

Curated telemetry and mission-data pipelines, feature store, model registry, evaluation harness, inference runtime and policy/approval gateway.
CPU/GPU/FPGA compute, deterministic message bus, time-series store, observability stack and isolated test environment.

Software, laboratory and analysis tools

Python, NumPy/SciPy, PyTorch or JAX, scikit-learn, notebooks and reproducible experiment tracking.
MLflow/model registry, Git, containers, CI/CD, telemetry dashboards, simulators and hardware-in-the-loop interfaces.

Verification and mission skills

Test distribution shift, drift, robustness, calibration, false-alarm/miss rates, latency, resource use and fail-safe behavior.
Maintain datasets, model cards, evaluation evidence, approval records, rollback controls and human-in-the-loop operating limits.

Authoritative technical references

The hiring sequence will follow operating milestones and project requirements. The position may begin through advisory, consulting or project-based work before moving into a full-time role.

Send Your Resume ↗