← Back to All Openings

Staff Systems Engineer – Federated Reputation Aggregation & Secure Aggregation (HRA)

corpora-jobs-1778796293285-db9d41c6 - Frontier Development
Systems EngineerStaff1 position
⚡

Why This Role is Different

Frontier Development Role

Architect and ship a resilient federated learning platform that blends anomaly detection, reputation scoring, and cryptographic aggregation to thwart poisoning attacks in real time.

The Frontier Element

This role implements the first sub‑linear, privacy‑preserving aggregation scheme that combines geometric anomaly detection with dynamic reputation vectors, a breakthrough for secure, large‑scale multi‑agent learning.

🔬

Project Context

Research Area

Hybrid Reputation Aggregation (HRA) for Federated Retraining

From: Adaptive Multi‑Agent Defense Against Adversarial Coordination

Why This Role is Critical

HRA is the linchpin that protects the shared model from poisoned or Byzantine updates while maintaining sub‑linear communication overhead, enabling RACE to scale to thousands of edge nodes.

What You Will Build

A production‑ready federated learning stack that fuses SHAP‑weighted Byzantine detection, momentum‑based reputation decay, and secure aggregation protocols (e.g., RAIN, homomorphic encryption) into the RACE engine.

🛠

Key Responsibilities

  • Design the HRA module: integrate SHAP‑weighted anomaly scores with a momentum‑based reputation vector that decays asymmetrically.
  • Implement secure aggregation using RAIN or equivalent protocols, ensuring end‑to‑end confidentiality of client updates.
  • Develop a lightweight, sub‑linear communication layer that scales to >10,000 edge devices without bottlenecks.
  • Create CI/CD pipelines for automated model retraining, validation, and rollback in the presence of detected anomalies.
  • Collaborate with the Applied Scientist to embed HRA outputs into the RACE trust‑aware communication layer.
🎯

Required Skills & Experience

Technical Must-Haves

Federated learning & secure aggregation

Expert
Designing protocols that preserve privacy while aggregating gradients from potentially malicious clients.

Differential privacy & homomorphic encryption

Advanced
Ensuring that client updates remain confidential during aggregation.

Anomaly detection (SHAP, geometric methods)

Proficient
Detecting poisoned or Byzantine updates in real time.

Distributed systems & network protocols

Expert
Building sub‑linear communication layers for large‑scale deployments.

Python, Go, Rust, Kubernetes, Argo Workflows

Expert
Implementing scalable, containerized pipelines.

Experience Requirements

  • 8+ years in distributed systems or federated learning.
  • Proven experience building secure aggregation or privacy‑preserving ML pipelines.
  • Track record of delivering production‑grade systems with sub‑linear scalability.

Education

PhD or Master’s in Computer Science, Electrical Engineering, or a related field with a focus on security or distributed systems.

⭐

Preferred Skills

  • Experience with RAIN or similar shuffle‑model protocols.
  • Knowledge of blockchain‑based federated learning frameworks (e.g., PQS‑BFL).
🤝

You Will Thrive Here If...

  • Owns end‑to‑end system design and shipping.
  • Comfortable making trade‑offs between security, performance, and scalability.
  • Excited by solving problems that have no textbook solutions.
📈

Impact & Growth

12-Month Impact

Deploy a fully automated HRA pipeline that achieves >98% model accuracy while keeping per‑client communication below 1 MB, enabling RACE to operate in a 10,000‑node UAV swarm within 12 months.

Growth Opportunity

Scale the platform to millions of nodes, lead cross‑domain security initiatives, and mentor a growing team of systems engineers.

Ready to Push the Boundaries?

If this sounds like the challenge you have been looking for, we want to hear from you. We value what you can build over where you have been.