- AgentPulse | AI Agents Directory
- Posts
- #27: Inside the AI Agent Boom: What Salesforce, Stanford, and KPMG Just Revealed
#27: Inside the AI Agent Boom: What Salesforce, Stanford, and KPMG Just Revealed
AgentPulse: Your Weekly Dose of AI Agents News.

Welcome back, AI Agent Enthusiast!
In today’s Agent Pulse:
📢 Top AI Agents News
⚔️ Agent Arena
Featured Agents
Free courses
Papers
📢 TOP AI Agents News
⚙️ AgentForce 3: Enterprise Agents Grow Up
Salesforce just rolled out AgentForce 3, and it’s no longer just about “trying agents” it’s about running your business with them.
What’s new:
Command Center gives teams full observability: monitor agent health, failures, escalations, and performance in real time.
Built-in support for the Model Context Protocol (MCP) lets AgentForce connect across 30+ services: Stripe, Notion, AWS, Slack without complex integrations.
Launches with 100+ prebuilt industry actions, lower-latency Atlas architecture, and global coverage for large orgs.
What it means:
Salesforce is pushing hard into enterprise-grade agent orchestration and it’s working. AgentForce has already been deployed across 8,000+ customers, cutting case handling time by 15% and achieving 70% autonomous resolutions in some use cases.
Takeaway:
This is no longer "pilot phase." Agent governance, observability, and trust are becoming standard for enterprise-scale deployment. If you're building agent infrastructure or targeting B2B workflows, this is your competitive bar.
📊 Stanford’s WORKBank: Where Workers Want Agents
Stanford researchers just dropped WORKBank, a massive study surveying 1,500 U.S. workers across 844 real-world job tasks, focused on how people actually want AI agents to help.
Key findings:
Workers are pro-agent but with limits. They’re all-in for automation of low-value, repetitive tasks, but wary of full autonomy in judgment-heavy roles.
The study segments tasks into four buckets:
- Green Light (automate it)
- Red Light (hands off)
- R&D (high complexity, high value)
- Low Priority (not worth automating)Collaboration is key, workers consistently preferred AI agents as copilots, not replacements.
🔍 Takeaway:
If you’re building agents for real-world users, don’t just chase capability align with human expectations. Augmentation wins over automation. Co-agents > auto-agents.

📈 KPMG Q2 2025 AI Survey: From Experiment to Execution
KPMG’s latest AI Pulse Survey confirms what many in the space already feel: the pilot era is over, and execs are pushing for results.
Key signals from Q2 2025:
33% of companies have deployed agents in production (up 3x from last year).
51% are building hybrid agent strategies, blending custom and prebuilt stacks.
82% of execs expect AI to reshape their competitive landscape in under 24 months.
But it’s not without friction:
69% cite data privacy as a growing concern (from 42% Q4 2024).
55% regulatory concerns (from 42%),
56% data quality (from 49%)
Takeaway:
AI agents are now a boardroom priority but only the solutions that are composable, trusted, and secure will make it to scale. Builders who solve for deployment pain will win.

Find out why 1M+ professionals read Superhuman AI daily.
In 2 years you will be working for AI
Or an AI will be working for you
Here's how you can future-proof yourself:
Join the Superhuman AI newsletter – read by 1M+ people at top companies
Master AI tools, tutorials, and news in just 3 minutes a day
Become 10X more productive using AI
Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.
AArena: All-in-One AI Platform
Bring your task. Get best results.
TOP 3 AI on June 29:
Llama 4 Maverick 17B
Llama 3.3 70B Instruct Turbo
Llama 3.3 70B Instruct
Other News Headlines
PwC (June 27, 2025): Launched “Agent OS,” an enterprise AI command-center that provides a unified framework to build, orchestrate, and integrate multiple AI agents and workflows at scale, helping organizations move beyond experimentation to full AI adoption (source)
Infinitus & Salesforce (June 25, 2025): Announced an expanded partnership so healthcare clients can invoke Infinitus’s voice AI agents via Salesforce Health and Life Sciences Clouds to automate complex administrative tasks (e.g. drug and benefits verifications) (source)
Andela & Emergence AI (June 25, 2025): Revealed a strategic partnership to train software engineers in multi-agent (“agentic”) AI systems, combining Emergence’s agentic platform with Andela’s global developer network to accelerate enterprise adoption of AI agent solutions (source)
Emergence AI (June 24, 2025): Unveiled CRAFT, a natural-language “agent crafting” platform for enterprise data pipelines. CRAFT lets business users describe goals in plain English and instantly spin up self-verifying multi-agent systems to automate complex data workflows (source)
Silverfort (June 27, 2025): Launched AI Agent Security, a new solution that treats each AI agent as an identity tied to a human owner. It enforces real-time access controls for agents (especially those using the Model Communication Protocol), preventing overreach or data leakage by applying human-like identity governance to autonomous agents (source)
Schroders & Google Cloud (June 25, 2025): Google Cloud announced that investment manager Schroders built a multi-agent financial research assistant using Vertex AI Agent Builder. The system uses specialized AI agents to autonomously gather data, perform analysis, and generate reports – dramatically speeding up equity research (source)
Skyfire (June 27, 2025): Introduced AI Agent Checkout, a payments technology specifically designed for transactions conducted by AI agents. The solution integrates payment and identity verification processes for agent-driven commerce scenarios (source)
kama.ai & Cohere (June 26, 2025): Announced that Cohere’s secure enterprise LLMs are now integrated into Kama.ai’s “Sober Second Mind” Hybrid AI Agent platform. This lets Kama.ai’s agents leverage Cohere’s generative models for creative tasks while maintaining Canadian data-governance and knowledge-graph accuracy (source)
Vertesia (June 26, 2025): Launched a new Autonomous Agent Builder, a low-code platform that enables business users to create and deploy AI agents. Agents built on this platform can autonomously plan and execute multistep business processes (ingesting documents, making decisions, etc.) with enterprise-grade tooling, security, and data-preparation (source)
PepsiCo (June 24, 2025): Announced it will deploy Salesforce’s Agentforce platform to bring autonomous AI agents into its operations (e.g. for customer support, sales, inventory management). PepsiCo is among the first major food-and-beverage companies to scale digital “robot workers” across its business, freeing human employees to focus on strategic tasks (source)
✨ Featured Agents
TeammatesAI: Autonomous AI Teammates
OraczenAI: Custom AI agents that simplify complex workflows
Twig: Automate Tier 1 support with AI Agents
TensorStax: Autonomous AI Agents for Data Engineering
TheLibrarian.io: WhatsApp AI Personal Assistant
Agentverse: Search and Discover AI Agents
🎓 FREE Courses
BeeAI: Agent Communication Protocol
Anthropic: AI Fluence Course, designed for everyday users of AI.
HuggingFace: Model Context Protocol (MCP)
Microsoft: Generative AI for Beginners
OpenAI: Advanced Prompt Engineering
🎓 Papers
How'd we do? |
Reach 11,000+ Readers:
Newsletter is read by VCs, founders, engineers, managers and tech professionals.