← Back to All Openings

Lead Research Engineer – Local Robustness Certification & Edge Sandbox

corpora-jobs-1778796293285-db9d41c6 - Frontier Development
Research EngineerLead1 position
⚡

Why This Role is Different

Frontier Development Role

You will pioneer the first end‑to‑end, formally verified LRC system that runs on 32‑bit microcontrollers, enabling agents to autonomously certify their local robustness and isolate malicious code before it propagates.

The Frontier Element

This role blends formal verification, randomized smoothing, and secure enclave design into a single, deployable package—an unprecedented combination for edge‑deployed MAS.

🔬

Project Context

Research Area

Local Robustness Certification (LRC) with Cryptographic Sandboxing

From: Communication Graph Vulnerability to Malicious Agents

Why This Role is Critical

LRC is the foundation for all other defense layers; it must be lightweight, formally verifiable, and enforceable on resource‑constrained agents.

What You Will Build

A certified LRC module that computes local robustness scores, exchanges succinct certificates, and enforces per‑agent MAC sandboxes to contain code injection.

🛠

Key Responsibilities

  • Design a lightweight LRC metric that balances tightness of certification with computational feasibility on embedded hardware.
  • Implement a 2‑bit certificate exchange protocol that encodes local robustness and integrity checks.
  • Develop a per‑agent cryptographic sandbox using MACs and secure elements to contain potential code injection.
  • Integrate LRC thresholds with local reconfiguration logic to trigger edge addition/removal.
  • Validate the entire stack on a fleet of edge devices under simulated attack scenarios.
🎯

Required Skills & Experience

Technical Must-Haves

Formal verification and randomized smoothing techniques

Expert
Proving robustness guarantees for LRC on deep neural networks.

Embedded systems programming (C/C++, Rust)

Expert
Implementing LRC and sandboxing on 32‑bit microcontrollers.

Cryptographic MACs and secure element integration

Expert
Designing per‑agent sandboxes that isolate malicious code.

Graph theory (degree, clustering, local metrics)

Advanced
Computing local robustness scores from neighborhood topology.

Real‑time systems and low‑latency communication

Advanced
Ensuring certificate exchange and reconfiguration decisions meet strict timing constraints.

Experience Requirements

  • 7+ years in formal methods or embedded security for safety‑critical systems.
  • Demonstrated experience building lightweight verification tools for neural networks.
  • Hands‑on work with secure enclaves or TPMs on edge devices.

Education

PhD in Computer Science, Electrical Engineering, or a related field with a focus on formal verification or embedded security.

⭐

Preferred Skills

  • Experience with graph neural network verification frameworks.
  • Knowledge of edge AI inference pipelines and model compression techniques.
  • Familiarity with GNN‑based attack detection.
🤝

You Will Thrive Here If...

  • Comfortable owning the entire stack from theory to firmware.
  • Shows a bias toward shipping working prototypes and iterating quickly.
  • Thrives in a high‑impact, low‑hand‑off environment.
📈

Impact & Growth

12-Month Impact

Within 12 months, deliver an LRC module that runs on 32‑bit microcontrollers, achieving a 95% reduction in successful code injection attempts while enabling autonomous reconfiguration in under 200 ms.

Growth Opportunity

Scale the LRC and sandbox technology into a comprehensive edge‑security platform, leading a team that expands the solution to multi‑tenant IoT deployments.

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.