The workload of a teacher can be incredibly demanding: beyond delivering lessons, there’s planning, administrative work, student support, and countless other tasks competing for time. All this can quickly become unmanageable.

As a Multiverse coach working with an education charity, Rachel O'Hehir sees this first-hand. “What every teacher has told me is just how demanding their role is,” Rachel explains. “They’re under a lot of pressure which can easily lead to burnout. So, removing the administrative burden is a no-brainer, allowing teachers to reinvest that time into what’s important.” 

Teachers lose 2-3 hours each week on manually preparing and reformatting slide decks, so it was no surprise that one teacher Rachel was coaching wanted to find a smarter way of working. That’s why she enlisted the help of AI to transform the way lessons are planned.

She built a new AI tool: a slide engine that automatically creates lesson decks, applies standard formatting templates, and pulls in the government-required literacy and vocabulary homework content.

“By inputting all the key resources needed for a lesson, teachers receive a slide deck that’s 80% ready for teaching,” Rachel explains, “and from there, they can easily tailor each deck to make sure it’s ready for different lessons and students.”

With the heavy-lifting done for them, the slide engine saves teachers at least an hour a week each – a massive cumulative saving of time across the organisation. It also enforces quality control, automatically integrating school and government requirements without creating extra work.

“At the minute, it’s the teacher’s department using the tool, but the plan is to roll it out to the wider organisation, and ultimately the wider federation of academies, based on how successful she and the team have found it.” A huge achievement for her and her fellow teachers.

For prospective Multiverse clients, being able to share confidential information is essential, so they can confidently discuss skills gaps in their business, how they’re currently using AI, and where they need the most support. They typically ask for a non-disclosure agreement (NDA) - something Sam and the Multiverse Legal team were ready to handle. 

But this was a slow process.

“When a customer needed an NDA,” Sam explains, “they would often share their own templated version. That required a manual legal review, and due to the volume, it would create a 1-3 day turnaround time.”

Sam and the team would spend an hour thoroughly reviewing each NDA. This didn't just eat up time, it meant teams had to wait longer to start conversations and get to work with clients. In short, the longer an NDA took to approve, the longer customers waited to start upskilling their teams.

Sam knew this needed to change, and saw an opportunity to connect two systems using AI: Ironclad, a contract management system, and Wordsmith, a legal AI tool that reviews contracts against a pre-submitted customer playbook. “The key for me was to get Ironclad and Wordsmith talking,” Sam says. “How do we get that file between one tool and the other? And how do I get it back?”

So, Sam got to work. To start with, he developed the NDA playbook with the Legal team, defining exactly what was acceptable and what needed flagging. “I uploaded the playbook to Wordsmith,” Sam says, “and turned it into an AI agent, which can review and edit NDAs, generate a summary of any potential areas to address, and populate our company details.” He then built an automated trigger in Ironclad, which sends the NDA to Wordsmith once uploaded, and back again once reviewed by the AI agent.

“The turnaround time dropped from 72 hours to five minutes”, Sam explains. “Now instead of fully reviewing NDAs ourselves, we’re reviewing the AI agent’s first pass at them.” This has freed up time for both the Legal and customer facing teams, so they can focus on areas where they can add deeper strategic value.

Beyond speeding up the process, Sam also reflected on how it’s inspired a cultural shift in the wider team. “It’s definitely motivated colleagues to lean more into automation. For example, teams across the business now go to the Ask Legal channel in Slack.” This AI-powered Slack chatbot uses an internal knowledge base to answer queries. “We just need to review and approve the output it shares.” Win-win for Legal, win-win for Multiverse.

Until recently, Gemma spent a large part of her day sorting through goods from corporate donations and listing them on eBay. As an eBay Administrator for Age UK, Gemma plays a key role in raising funds for the charity’s work including services like its Telephone Friendship Service, free and confidential Advice Line and The Silver Line Helpline.

“We work with companies like Amazon that donate large quantities of items,” Gemma explains. “It can be anything, from skincare products to furniture.” All profits go directly back to Age UK’s important work.

But keeping up with demand was a grind. Between researching unfamiliar products, crafting accurate descriptions, and finding the right keywords to drive visibility, listing 25-30 items could take up to two and a half hours per day on average - taking away from her other duties managing the donations warehouse.

The variety of donations made things even trickier. “I’m not an expert on many of the items listed, so I would often have to frantically Google to make sure descriptions are accurate,” Gemma says. Specialist items were a particular challenge: “If it’s something like a car part, that takes a lot more time - I’ve got to make sure it’s the right thing.”

So Gemma decided to build a smarter way of working, leveraging her new AI skills. Using Microsoft Copilot to craft her prompts, she set up a ChatGPT Enterprise tool and fed it eBay’s best practice and tone of voice guidelines, giving every listing the best possible chance of a sale. “Copilot helped me write the prompt,” she explains, “and then I went through about three or four iterations before I got to the one I use now.”

The workflow is now remarkably simple. Gemma submits an item’s barcode to the tool, and within seconds ChatGPT pulls product information from across the internet and generates a polished title, description, and recommended price. It doesn’t matter how obscure the product is - the output is consistently accurate, and Gemma only needs a quick check before each listing goes live. “The stuff that I won’t know about, the internet will,” she jokes. “So it’s definitely made it easier.”

The impact has been significant. Time spent researching and writing each listing has dropped from 10-15 minutes to around five - and that freed-up time goes straight back into managing the warehouse. The team has also been able to add five more listings per day, boosting their inventory and the charity’s prominence on the platform. “The more things you have listed, the more sales you get,” Gemma notes. “Essentially, it bumps you up in visibility on eBay.”

Better listings have also reduced the number of customer questions Gemma has to field. When product descriptions are clear and detailed, buyers already have the information they need - another small win that adds up over the course of a busy day.

For Gemma, the bottom line is straightforward. AI-optimised listings are driving more revenue for Age UK, and giving her valuable time back for other work. “In the long run,” she concludes, “it is going to make more money for our department, the charity, and ultimately help the people we support.”