Charm Security - Applied AI Researcher

Charm Security is building the first agentic AI workforce for scams and human-centric fraud. Our AI agents work alongside fraud teams at Tier 1 banks in North America: investigating cases, guiding frontline teams through high-risk interactions as they happen, and gathering intelligence beyond the organization's perimeter.We do it by understanding manipulation, intent, and human behavior, through our proprietary HVE™ (Human Vulnerabilities and Exploits) framework, which combines fraud expertise with behavioral psychology. That's how we take institutions from detection to action.We're looking for an Applied AI Researcher to build the agents and models behind that platform. This role suits a data scientist or AI researcher who wants their research to ship: someone energized by deploying AI at scale, working under real-time and adversarial constraints, and seeing the impact on people who would otherwise have lost their savings.What You'll DoDesign, build, and evaluate the AI agents at the core of our platform, across fraud investigation, frontline guidance, and threat intelligence.Model how manipulation actually works: adapt and fine-tune LLMs to recognize intent, deception, and coercion in real conversations.Build scalable, real-time pipelines that turn behavioral signals, case data, and the alerts our customers already generate into decisions our agents can act on.Optimize AI models for latency, accuracy, and adaptability in high-volume fraud prevention use cases.Work AI-natively. Use coding agents and AI tooling to prototype, experiment, and ship faster.Stay at the forefront of LLMs, ML, and AI research, implementing the latest advancements in those fields.