During a recent fireside chat, Mimica CEO Tuhin Chakraborty and Head of Strategic Change at Cantor Fitzgerald Stephen Karas discussed how Mimica has quickly identified automation opportunities across the organization. The case study below is based on their discussion. To watch the full video click here. 

Replacing manual methods of automation discovery to unlock scale

A few years ago Stephen Karas, Head of Strategic Change at Cantor Fitzgerald, had a clear goal in mind: help Cantor Fitzgerald identify and act on automation opportunities across the firm. 

“It was born out of a desire to change our firm’s culture in terms of how we can make work more enjoyable, how we can make people’s jobs more satisfying, and actually help them work on things they want to work on,” he said.

He had no tools at his disposal, so he did what transformation leaders have done for decades: walk the floor.

Karas spent months moving through the back office: finance, risk, treasury, tax. He sat with teams and asked a simple question: what do you dislike about your job? In those conversations he started to uncover unfiltered pain points. From those conversations, 120 ideas emerged.

It worked. But it was slow. Every insight required his presence and someone else’s time. The process was hard to scale, and harder to sustain.

Mimica immediately impresses the C-suite

Karas was aware of tools that claimed to help the automation and transformation process. Years earlier, at another firm, he’d tried a recording-based tool. The output was what he now calls “a plate of spaghetti.” 

He was left to decipher a report that purported to show a map of the work being done by his teams and how the processes connected, but in reality was a dense tangle of lines on paper. There was so much overlapping data and branches that it was difficult to take action on the results.

Useful in concept, useless in practice. When he discovered Mimica he was excited, but skeptical.

Mimica suggested starting with a focused POC: record eight people on the London brokerage receivable team for a week, then review the output together. Karas agreed. 

The outcome was even more powerful than he could have imagined. Miner provided a list of all the tasks the team performed while Mapper provided a comprehensive process map showing all the variations.

 “We went through the process, and we were astounded at how the information came back in a way that was suitable for executive management.”

Not only was Karas impressed, but the executive team immediately grasped the takeaways. “Our CFO and COO were blown away in that executive meeting.”

Change management sets the tone for success

Once the program went live across the company, the questions started. What exactly does Mimica record? Can employees turn it off? Do they have any control? In financial services, those questions carry extra weight. HR, legal, and compliance teams all needed answers before the program could expand. Because Mimica has industry leading privacy controls, Karas felt confident the team would get on board.

He held team meetings, walked people through the what and the why. “We want to be better. We don’t want people doing inefficient things anymore,” he explained in those meetings. This upfront work helped bridge the gap between the technical outcomes and the human, emotional side. 

Even with initial buy-in, individuals were still naturally curious how the review process would work. Teams came in uncertain of what they’d see, unsure how their work would look when laid bare. Then something unexpected happened.

“I saw a meeting happen with people who were somewhat anxious, but it turned from an anxious feeling into, ‘Wow, people care about what I do. Let me tell you what I was actually doing in the output from Mimica.’”

Mimica gave employees something rare: data about their own work, presented back to them as evidence of its complexity and value. 

From team anxiety to advocacy

Six months later, the dynamic reversed entirely.

Team leads who were unsure about how much value they’d receive now see how Mimica can help accelerate their business initiatives; they no longer wait to be approached, they’re requesting licenses.

For the day-to-day users who had their work recorded, they’ve now seen how this process can remove unnecessary, tedious work that only served to slow them down. 

In a world where AI is moving fast, recording yourself isn’t a threat. It’s a way to get ahead of change rather than be swept along by it.

Karas now encourages the team not just to participate, but to lead. Mimica’s Analyst, an AI feature that allows teams to query their recorded data conversationally, “is a game changer,“ he says. “You can ask simple things like: give me automation opportunities, tell me for this process what’s the best-in-class tool and application, tell me how I should make the changes." 

Mimica has turned into a tool any employee can use to not only improve their day-to-day workflow, but it can also serve as an opportunity to be a leader for change and learn new skills related to orchestrating and managing AI. “Be the person of change. Don’t let the change happen to you,” Karas said.

The culture Karas describes is a long way from those first anxious meetings. The question staff are asking now is a different one entirely.

“It’s less anxiety and more of: can I get in line?”

How Mimica redesigned and automated operational workflows

Nowhere is this clearer than in the equity operations team.

The team processes trade allocations. Hundreds of client emails arrive every day, each in a slightly different format, each requiring manual navigation across multiple systems to validate and take action on. The team knew they were inefficient. What they lacked was data to prove it. Mimica gave them that: the process ranked as the second most inefficient task across the entire operation.

From there, the full Mimica workflow followed. Miner captured the scope of the work. Mapper documented the as-is process and modeled a to-be state, with Mimica’s team facilitating the redesign session. 

The result was a new automated pipeline: an OCR system reads incoming emails, AI validates the key data fields — client name, trade account number, equity sold — with confidence scores, and a human analyst reviews and approves each trade before it enters the underlying transactional system.

The big takeaway? The team didn’t see this as an overreach or critique of how they were doing their job. It was a relief. “The team was very excited about the process because they’d lived with this for so long. Ultimately, they’re now in their better world,” said Karas.

Using Mimica’s findings to uncover more automation opportunities

The equity operations automation is one of 55 projects Cantor Fitzgerald completed in six months with Mimica. All managed, end to end, by a single person handling task identification and process reengineering. The team is always on the lookout for new ways to improve their processes and take their work to the next level. 

One recent addition is accelerating things further. The Mimica Analyst has boosted the development process. 

Karas’s developers use the tool to generate development prompts, and take those prompts into tools like Cursor or Claude Code to start building automations — shaving off time those developers would previously have spent explaining context to the LLM.

"Using AI to create AI has enormous potential."

Transformation is a constant process and Mimica is the foundation

Karas thinks about AI adoption as a three-step process: 

1. Enable your people: Enabling individuals on AI tools like ChatGPT or Claude that they use in their day-to-day  
2. Implement in your systems: Ensuring third-party AI applications are vetted and implemented correctly to produce the best results
3. Automate your processes: Discovering where AI can revolutionize and automate processes and systems that are already in place 

Mimica, he said, is the foundation of step three. Process intelligence delivered by Mimica determines which processes are worth transforming before you commit to transforming them — and it reveals how to transform them in the right way.

For organizations considering where to start, his advice is simple: start with one project. That’s where he was convinced. That’s where the skepticism fell away.

“That was the ‘aha’ — yes, it’s here, it’s ready. And we’ve been off to the races ever since.”