With Credal, Lattice deployed impactful AI across HR, IT, and Security

Founded in 2015, Lattice is a people-management platform that helps companies run performance reviews, one-on-ones, and engagement programs. As Lattice scaled, internal teams like HR/People Ops, IT, and Security were spending a growing share of their day answering the same employee and customer questions, often across Slack, Google Docs, and ticketing tools.

At peak, these teams were collectively handling hundreds of repetitive questions per week, leading to slow response times, constant context switching, and limited time for higher-impact work.

After its DIY experiment, Lattice developed clear requirements for any enterprise AI solution.

First, it had to be Slack-first. Employees already asked questions in Slack, and over 70% of internal support requests originated there, so any AI solution needed to meet them there.

Second, it needed to draw from trusted internal sources, including internal FAQs, policies, Slack history, Salesforce data, and procurement systems like Zip.

Third, security was vital. Lattice already had a strong permission model in Google Workspace, and the AI had to respect those boundaries and not introduce new risks.

Finally, the system needed to be usable by all employees, including non-technical folks. People Ops, Product, and Sales had to be able to develop and use agents without constantly relying on engineering.

Credal stood out because it could publish AI agents directly into public and private Slack channels, connect to Lattice’s existing systems, and respect existing permissions seamlessly.

People Ops: 24/7 answers to employee questions‍

Lattice's People team fields constant questions about benefits, policies, and "how things work here." They built a People Ops assistant trained on internal FAQs and documentation so employees could ask questions any time without waiting for a human reply.

Within the first few weeks, the assistant was handling 60–70% of common People Ops questions autonomously, reducing average response time from hours to seconds.

Instead of a People Ops teammate digging through docs or past messages to answer “What’s our policy on X?”, the agent provided answers immediately. A human only stepped in when necessary, saving the team an estimated 10+ hours per week.

Security: Faster answers to customer questionnaires

Lattice’s Security team is small, and Sales frequently brought them the same customer questions, often copied from last week’s deal cycle. Lattice built a Security assistant so Sales could get consistent, accurate answers faster.

This reduced turnaround time for security responses by 50–60%, helping deals move forward without waiting on a limited security team.

‍IT support in Slack

After the People and Security assistants worked, Lattice built an IT agent, using Credal to handle first responses to common internal support questions so that IT could focus on escalations and deeper work.

The IT assistant now deflects 30–40% of routine IT inquiries, and first-response time dropped from minutes or hours to near-instant.

Procurement Status

Lattice also piloted an agent to answer procurement status questions. Employees exported request data from Zip and connected it to an assistant so employees could ask, “What stage is my software request in?” and get an answer in seconds, without clicking through Zip. This eliminated dozens of manual status checks per month and removed a common source of internal back-and-forth between employees, IT, and Finance.

Lattice’s early metric was simple: reduce manual response time. As adoption spread across teams, they evolved that into a broader business metric: time-to-answer across the company.

Across HR, IT, and Security:

• Average internal response time dropped by more than 50%
• Thousands of questions per month are now answered directly in Slack by AI agents
• 30–40% of internal support questions no longer require human intervention
• 300+ hours are saved per month across teams through reduced manual work

Just as important, publishing agents directly into Slack reduced the change-management burden: employees didn’t have to learn a new workflow, driving high adoption with minimal training.

Lattice had a very positive experience working with Credal to create useful internal agents employees could actually rely on day to day. Credal helped Lattice deploy agents in Slack, connect knowledge with sources like Google Workspace, and keep the existing permission model intact. Lattice also brought compliance into the rollout early and Credal’s default protections were able to reduce the chance of sensitive data leakage. Jeter explained that Lattice has now been able to make a “paradigm shift from operating reactively to operating more proactively.”

Have questions on the above or want to know how these apply these use cases more specifically to your needs? We’re happy to chat. Drop us a line at sales@credal.ai!