Job Description
Note: The job is a remote job and is open to candidates in USA. Interface.ai is a company focused on reimagining banking through innovative AI solutions. They are seeking a Staff AI Engineer specializing in Human-Computer Interaction and Cognitive Systems to lead the development of multimodal interactive agents that enhance customer experience in financial services. Responsibilities • Design and implement perception pipelines that combine text, visuals, and UI semantics for agent grounding and decision-making • Build systems that allow agents to understand and interact with software UIs (browser DOMs, screenshots, or structured layouts) like a human operator • Develop planning and reasoning modules enabling multi-step task execution, contextual memory, and human-in-the-loop collaboration • Integrate LLMs and multimodal models for adaptive, goal-oriented behavior using techniques like ReAct, Tree-of-Thought, or Hierarchical Planning • Architect agent behaviors around transparency, safety, and trust — ensuring every AI decision or action is explainable and controllable • Collaborate with product, UX, and cognitive researchers to design experiences that feel intuitive, reliable, and emotionally intelligent • Implement safe sandbox environments for browser or desktop interaction (Firecracker/gVisor-based isolation) • Build reinforcement and feedback loops for continuous learning and evaluation of agent performance • Partner with Bot Platform, AI Infrastructure, and Compliance teams to ensure that cognitive systems scale securely and responsibly • Mentor engineers and applied scientists in agent design, multimodal integration, and AI safety Skills • 10+ years of experience in software, AI systems, or cognitive computing, with at least 2+ years building multimodal or interactive AI applications • Advanced proficiency in Python (PyTorch, JAX, TensorFlow) and at least one programming language (Go or Node.js) • Expertise in LLMs, computer vision, or multimodal architectures (e.g., CLIP, BLIP, Flamingo, GPT-4V, Gemini) • Deep understanding of human-computer interaction principles, cognitive modeling, and user-adaptive AI • Proven experience integrating LLM-based agents with external tools or UIs (browser automation, API control, or RPA) • Experience designing or evaluating planning and reasoning agents (e.g., ReAct, AutoGPT, OpenDevin, Voyager) • Familiarity with reinforcement learning, behavior cloning, or imitation learning in simulated environments • Strong background in observability, safety, and interpretability of AI systems • Excellent communication and collaboration skills — able to translate between research and product engineering • Advanced degree in Computer Science, AI, Cognitive Science, or Human-Computer Interaction Benefits • 100% paid health, dental & vision care • 401(k) match & financial wellness perks • Discretionary PTO + paid parental leave • Mental health, wellness & family benefits Company Overview • interface.ai is an award-winning out-of-the-box Intelligent Virtual Assistant(IVA) for Banks & Credit Unions. It was founded in 2019, and is headquartered in San Francisco, California, USA, with a workforce of 51-200 employees. Its website is Company H1B Sponsorship • interface.ai has a track record of offering H1B sponsorships, with 3 in 2025. Please note that this does not guarantee sponsorship for this specific role. Apply tot his job