“Credal gives us the autonomy to action and gate systems in ways I haven’t seen other tools match… Being able to set up a dedicated RevOps versus broader org MCP, that’s the game changer. I am constantly updating Salesforce mass record updates using our RevOps MCP. I cannot live without it.”
Guidewheel is an AI-powered FactoryOps platform built for the manufacturing sector, using sensors and cloud analytics to give teams real-time visibility into equipment performance. By detecting operational patterns and anomalies, Guidewheel enables hundreds of manufacturers worldwide to reduce downtime, optimize throughput, and boost energy efficiency.
When Guidewheel's executive leadership began pushing hard for company-wide AI adoption, it was difficult for the Revenue Operations team to scale AI usage safely.
As Callum Batten, Guidewheel's Sr. Revenue Operations Manager explained, "We needed some way to securely access our data that wasn't just giving ungated API access. I don't want Claude to be able to go in and query our Salesforce whenever it wants and have access to it. There are contracts with our customer data that we have said we won't do that."
Compounding this challenge was Guidewheel's tech stack: the company had no full-copy Salesforce sandbox and no dedicated data backup tool. A rogue AI write action could corrupt live CRM data.
At the same time, employees eager to use tools like Claude were querying Salesforce data on their own. Because non-technical team members didn't know complex CRM schemas or RevOps filtering rules, they were generating a lot of slop.
“People would come to us and be like, ‘Can you validate this?’ And I was like, ‘What data did you pull? How did you pull it? What were the filters you used? What did you ask?’ Without them knowing Salesforce, there’s no way they’re going to have all the filters of what we care about… You’re basically doing a scavenger hunt, which is not the best time for anybody.”
Guidewheel brought in Credal as their centralized AI platform and MCP gateway to solve the problem of scaling AI safely: the RevOps domain expert can centrally configure security, business context definitions, and permission mirroring once, while delegating AI and data access to users across the company.
Deploying AI usually has a tradeoff: give everyone full access and risk your data, or lock everything down and slow your team down. Guidewheel used Credal to create role-based MCP servers over their business context, acting as "bumpers on a bowling alley lane" so employees could use AI safely without touching critical backend logic.
The RevOps MCP (full write): Configured with full read and write capabilities, this environment allows RevOps leads to execute mass-record updates in Salesforce in natural language.
The Organizational MCP (read only): Cloned for the rest of the organization, this environment strips out write permissions and restricts visibility to certain fields. Users can query account data without any risk of overwriting production records.
Guidewheel created a field directory for their business context that auto-updates in Credal. As Callum shares, "Credal natively has access and understanding of our objects and our structure in Salesforce. Regardless of whether someone is querying via the Credal web app, the chat agent, or in Claude, we don't need to worry, did they download the right skill to have the content? We just know we have the faith and comfortability that it's set up correctly on our end."
With Credal in the picture, RevOps shifted from ticket-solvers into high-velocity internal product builders. Using Credal as the connective brain across their tech stack, RevOps engineered several high-impact internal solutions:
"The War Room" App (Credal + Claude + GitHub + Railway): A custom application that automated complex daily workflows for sales executives, saving 4 hours a day of manual work.
Automated Data Backup (Credal + Postgres + Railway): An automated Salesforce data backup pipeline, giving Guidewheel a safety net at zero additional software cost.
Credal became so deeply woven into Guidewheel's daily operations that executives rely on it constantly. As Callum noted: "Everybody knows what Credal is. People freak out when the Credal MCP disconnects for whatever reason… like the world is ending."
Guidewheel evaluated using native Claude skills or plugins, but they quickly realized they placed a huge mental load on end users. Expecting busy account executives to download local GitHub repositories and remember which skill to call resulted in very low adoption.
"Even with skills, you are assuming that the person knows to do the right thing at the right time and call the right skill. Credal takes the mental load of understanding system architecture off," Callum explained. Furthermore, native skills didn't offer access controls or central audit logging needed to enforce customer data contracts, both of which Credal provided.
While building a basic MCP connection in-house is doable, in an ecosystem running multiple interconnected go-to-market systems (Salesforce, Gong, HubSpot, and Zapier) in-house maintenance scales non-linearly. Every schema change, MCP protocol update, and new internal workflow requires continuous engineering oversight. Guidewheel determined that buying something out-of-the-box was a better return on their investment.
Although most people used Claude, there were still many users who preferred ChatGPT or Codex. Guidewheel wanted an architecture that was agnostic to model providers to meet users where they were. Decoupling their agentic workflows from the model layer allows every query to be grounded in the exact same source of truth, regardless of which chat interface is used.
Reach out to sales@credal.ai to see how Credal's MCP can bring the same governed, accurate AI experience to your team.
One platform for all agents.
Full visibility for admins, full access for teams.
