ECNETNews reports that Salesforce has introduced two groundbreaking AI models—xGen-Sales and xLAM—aimed at enhancing its Agentforce platform, which synchronizes human and autonomous AI agents to elevate business efficiency.
xGen-Sales is a specialized AI model tailored for sales-related functions, including the generation of customer insights, call summaries, and pipeline management. By automating routine sales tasks, this model allows sales agents to concentrate on strategic initiatives. With its capabilities, xGen-Sales significantly enhances Agentforce’s efficiency in managing customer interactions, nurturing sales pipelines, and offering timely assistance to sales teams.
The xLAM (Large Action Model) series introduces AI models crafted to execute multifaceted tasks and initiate actions within business systems. Unlike conventional LLMs (Large Language Models) that primarily focus on content, xLAM models excel in function-calling—enabling AI agents to operate autonomously, from initiating workflows to processing data independently. These models vary in size and functionality, offering everything from compact, on-device versions to larger models suited for industrial applications.
Developed by Salesforce AI Research, the xLAM models utilize advanced data-generation technology to enhance performance. Preliminary results show that early iterations of xLAM models have outperformed existing large models on critical benchmarks. Notably, the xLAM-8x22B model claimed the top spot in function-calling tasks on key leaderboards, exceeding larger models like GPT-4.
This latest innovation empowers businesses to scale their AI-driven workflows more efficiently, enabling organizations to automate complex tasks, refine sales strategies, and improve resource allocation.
The non-commercial xLAM models are set for community evaluation, while proprietary versions will enhance Agentforce. The xGen-Sales model has successfully completed its pilot phase and is poised for widespread release.
