Everyday AI Podcast ā An AI and ChatGPT Podcast Purpose-Built Enterprise AI Agents: What Actually Works
38 snips
Jan 8, 2026 
Prashanthi Padmanabhan, VP of Engineering at LinkedIn, specializes in building AI agents for recruiting. She discusses why narrow, purpose-built agents outperform general ones in enterprise settings. Prashanthi explains how to reverse engineer recruiter workflows to define agents' roles and highlights the importance of human-in-the-loop design in enhancing efficiency. She also shares insights on using large language models to uncover talent and the necessity of trust through transparency. Metrics indicate significant improvements post-assistant implementation, marking a shift in AI's role within recruiting.
AI Snips
Chapters
Transcript
Episode notes
Narrow Purpose Beats General Agents
- Agents today succeed when built for a narrow, well-defined purpose rather than as a general all-purpose worker.
- Purpose-built agents deliver measurable goals by focusing on a specific workflow and data set.
How Hiring Assistant Targets Recruiter Pain
- Prashanthi described building LinkedIn Hiring Assistant to handle recruiting workflows like role definition, sourcing, and outreach.
- The agent mines large candidate sets so recruiters can focus on human-centric tasks like rapport and strategic hiring.
Build With Customers, Not In A Boardroom
- Work directly with customers from day one and iterate rapidly on product, models, and UX.
- Shift from pure asynchronous automation to a conversational, co-working experience where recruiter and agent tag-team.