AI agents and bots
Assistants for sales, support and back-office teams that answer from your data, take action and hand off to a person when it matters.
e.g. triage incoming emails, draft replies, queue them for approval
Learn more
Claude · AI agents · Salesforce · Automation
AI agents thatdo the work.
Connect Salesforce, your inbox and your documents. Automate everyday work, with you in control.
Salesforce & CRMAgentforce, Data 360 and managed admin and development.
AI agents & automationClaude agents for sales, support, operations and marketing.
Cloud & integrationAgents in your own AWS, Azure or Google Cloud, connected to everything.
- 01DiscoverYour goals and the best first use case.
- 02BuildAgents designed for your workflows.
- 03ReviewTested, refined, and you approve.
- 04LaunchDeploy and scale across your team.
Works with the tools you already use, and almost anything with an API
Tap a category to see examples.
Try it yourself
Pick a task and watch the agent work. It's a simulation with sample data, so nothing real is touched. Browse 39 more ideas →
Conversation
Tool calls via MCP
Tool calls appear here as the agent works.
A scripted preview with made-up records. Real agents run on your data, with your permissions, and every action is logged.
What we build
Agents that connect to your real systems and stay inside the limits you set.
MCP servers
Model Context Protocol servers that give Claude safe, logged access to your internal tools, databases and APIs, and nothing more.
e.g. read-only access to orders, write access only after approval
Learn more
AI for Salesforce
AI inside Salesforce: record summaries, next-step suggestions, data clean-up and automatic logging, built by Salesforce developers.
e.g. flag deals with no activity in 21 days and draft follow-ups
Learn more
Back-office automation
Invoices, expense checks, onboarding paperwork and weekly reports: the repetitive checking that eats your team's week.
e.g. match invoices to purchase orders, flag the ones that don't fit
Learn more
Managed Salesforce
Your Salesforce, run by our team.
Admin and development, built, tested and documented. Save 60% or more compared with a full-time hire.
Free health check · First two weeks free
See what you'd save →Compared with the full cost of a full-time hire, including salary and benefits.
AI for marketing
Come with me. I’ll show you how it runs itself.
Bots plan, write, post, reply and report. You just approve.
Take me there →How we work
Small first, then real. You see a working agent before you commit to a rollout.
Pick one workflow
We choose a single repetitive task with clear value and a result you can measure.
Prototype and test
A working agent on sample data first, then on a safe slice of your real data.
Deploy and hand over
Production rollout with approval steps, logging and documentation your team can run.
Questions
Something else? Email hello@nodepointai.com.
What is MCP?
The Model Context Protocol is an open standard that lets an AI model such as Claude use tools and data sources through a common interface. We build MCP servers so an agent can reach your systems with only the access it needs.
Do you only work with Salesforce?
No. Salesforce is one of the systems we connect. We also build agents for back-office work such as invoices, expense checks, onboarding paperwork and reporting.
Will an agent act without permission?
Not for anything that matters. Payments, emails to customers and record changes wait for a person to approve, and every tool call is logged.
How does a project start?
With one workflow and a fixed-scope pilot. We prototype on sample data, test on a safe slice of your real data, then deploy with logging and documentation.
Start a conversation
What repetitive work would you hand to an agent?
hello@nodepointai.com