Applied Scientist

About Dialpad

Dialpad is the AI-native business communications platform. We unify calling, messaging, meetings, and contact center on a single platform - powered by AI that understands every conversation in real time. More than 70,000 companies around the globe rely on Dialpad to build stronger customer connections using real-time, AI-driven insights. We’re now leading the shift to Agentic AI: intelligent agents that don’t just analyze conversations but take action by automating workflows, resolving customer issues, and accelerating revenue in real time. Our DAART initiative (Dialpad Agentic AI in Real Time) is redefining what a communications platform can do.

Your role

As an Applied Scientist at Dialpad, you'll be a key driver within our AI team, conducting R&D to power the next generation of autonomous voice agents. While traditional NLP focuses on analyzing static text, your work will center on real-time, multimodal systems that can listen, reason, and take action during live customer interactions. A major focus of the team is advancing DialpadGPT, our proprietary LLM, specifically optimizing it to drive orchestrated agentic workflows and handle complex task execution. Beyond the technical skills, we are a team that values collaboration, continuous learning, and the application of diverse perspectives to solve complex problems. Collaboration will be key as you work alongside our engineering, design, and product teams to build groundbreaking agentic applications. If you're passionate about agentic and multimodal AI and contributing to a team that's changing the face of business communications, you'll find yourself right at home with us.

This position reports to the Senior Manager of the NLP team and has the opportunity to be based in Vancouver, BC.

What you’ll do

• Research and develop state-of-the-art algorithms for autonomous voice agents, specifically focusing on real-time speech processing and reasoning loops.
• Advance DialpadGPT: Design and execute distributed training strategies to optimize our proprietary LLMs for agentic behaviors, including precise tool use, instruction following, and latency-constrained generation.
• Conduct rigorous evaluation and monitoring of model performances and troubleshoot issues with a keen understanding of resultant business impacts.
• Design and implement orchestration layers that effectively chain LLMs with external tools and APIs to solve complex customer problems autonomously.
• Work with large-scale multimodal datasets (text, audio) to improve model robustness and alignment.
• Collaborate with engineering, product, and design teams to deploy scalable, low-latency models and algorithms in production.
• Submit papers to top-tier academic conferences (ACL, EMNLP, NeurIPS) and contribute to the team’s research culture.

Skills you’ll bring

• Master’s or PhD degree in Computer Science, Machine Learning, Computational Linguistics, or a related quantitative field.
• 2+ years of industry experience in Machine Learning/NLP for Master’s degree holders, or 1+ years for PhD holders.
• Deep understanding of LLMs: Demonstrated experience with training, fine-tuning (PEFT/LoRA), and alignment techniques (RLHF/DPO) for specific domains or tasks.
• Experience with Agentic Systems: Familiarity with building autonomous agents, including concepts like tool use, function calling, reasoning chains (CoT), and memory management.
• Strong proficiency in Python and PyTorch, with the ability to write clean, production-ready research code.
• Research Track Record: A history of publishing in top-tier conferences (ACL, EMNLP, NeurIPS, ICASSP) is highly valued.
• Multimodal Awareness: Familiarity with speech technologies (ASR, TTS) or processing real-time audio streams is a strong plus.
• Ability to bridge the gap between research and product, translating complex technical concepts into business value.
• Familiarity with version control tools like Git for collaborative projects.