Lead Applied Scientist – Trust‑Aware Sensor Fusion & LLM Robustness (TASF‑DFOV + RS‑LLM‑MAS)

corpora-jobs-1778796293285-db9d41c6 - Frontier Development

Applied ScientistLead1 position

⚡

Why This Role is Different

Frontier Development Role

Lead the creation of a trust‑aware perception engine and a statistically‑certified LLM smoothing layer, ensuring that every sensor reading and language output is vetted for integrity before influencing collective decisions.

The Frontier Element

This role fuses Dirichlet‑based trust distributions with ray‑tracing‑derived dynamic FOV, and introduces the first randomized smoothing scheme for LLM agents that provides a certified radius against adversarial hallucinations.

🔬

Project Context

Research Area

Trust‑Aware Sensor Fusion with Dynamic Field‑of‑View (TASF‑DFOV) and Randomized Smoothing for LLM‑Based MAS (RS‑LLM‑MAS)

From: Adaptive Multi‑Agent Defense Against Adversarial Coordination

Why This Role is Critical

These two modules provide the runtime assurance and perception robustness that make RACE safe for autonomous UAVs, cyber‑physical networks, and finance systems.

What You Will Build

A Bayesian HMM‑based fusion engine that weights LiDAR, vision, and radio data by dynamic trust PDFs, coupled with a randomized attention‑masking layer that bounds LLM hallucinations within the MPAC coordination protocol.

🛠

Key Responsibilities

Design the HMM‑based trust‑aware fusion engine, incorporating dynamic FOV estimation via point‑cloud ray‑tracing and Dirichlet trust PDFs.
Implement the randomized attention‑masking mechanism for LLM agents, integrating it into the MPAC multi‑principal protocol.
Develop end‑to‑end validation suites that measure attack detection rates, localization error, and hallucination influence bounds.
Collaborate with the DRAT engineer to ensure that sensor and language outputs are aligned with the dynamic role assignment.
Publish formal proofs of bounded influence and provide open‑source benchmarks for the community.

🎯

Required Skills & Experience

Technical Must-Haves

Bayesian inference & Hidden Markov Models

Expert

Modeling trust as a latent state and updating it in real time.

Sensor fusion (LiDAR, vision, radio)

Advanced

Implementing dynamic field‑of‑view estimation and trust‑weighted fusion.

Large‑language‑model safety & randomized smoothing

Proficient

Bounding hallucination influence in multi‑agent coordination.

MPAC multi‑principal coordination protocols

Expert

Ensuring robust message exchange among principals.

C++, Python, ROS2, NVIDIA Isaac SDK

Expert

Deploying perception pipelines on edge hardware.

Experience Requirements

8+ years in robotics or autonomous systems with a focus on perception or LLM safety.
Published work on trust‑aware fusion or randomized smoothing for language models.
Experience deploying real‑time perception stacks on UAVs or autonomous vehicles.

Education

PhD in Robotics, Computer Science, or a related field with a strong emphasis on perception or AI safety.

⭐

Preferred Skills

Experience with autonomous traffic control or smart‑city testbeds.
Knowledge of control‑barrier functions and event‑triggered safety.

🤝

You Will Thrive Here If...

Owns the entire product lifecycle from research to deployment.
Thrives in environments where safety and interpretability are paramount.
Enjoys turning complex theoretical guarantees into tangible, real‑world systems.

📈

Impact & Growth

12-Month Impact

Deploy a trust‑aware fusion engine that detects >95% of spoofing/jamming attacks while keeping localization error below 0.8 m, and a randomized smoothing layer that limits LLM hallucination influence to a certified radius, all within 12 months.

Growth Opportunity

Lead cross‑domain safety initiatives, expand the trust‑aware framework to decentralized finance and IoT, and mentor a team of applied scientists.

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.