Categories: EdTech
Mathema is one of the largest EdTech platforms for online math tutoring. Every month, more than 12,000 students from various countries attend over 55,000 one-on-one lessons with 1,000+ tutors.
This scale means tens of thousands of support requests each month from students, parents, and teachers.
To maintain high-quality service amid this growth, Mathema’s team turned to automation — and it paid off. Today, a significant portion of requests are successfully handled by AI, without compromising the human touch.
The Challenge
2022
By 2022, Mathema already had more than 1,000 students and over 100 tutors. Requests were coming in from all channels: chat widget, Telegram, Viber, email, and social media.
Things got even more complicated as Mathema prepared to launch in new markets. They needed to rapidly scale support operations across new languages and time zones.
“Sometimes parents would write in the middle of the night, a tutor couldn’t find a link, and we were trying to handle everything manually. We knew: either we automate or we drown.”
Maryna Zaitseva, Head of Support & Operations at Mathema
This added pressure on the team and required an even more flexible approach to handling inquiries.
2025
In 2025, the company faced a new challenge: rapid growth of its student and tutor base across several markets and languages began to exponentially increase the volume of lessons and support requests.
For comparison: from 6,000 requests per month in 2022 to up to 30,000 in 2025.
There were only two options: either scale the support team proportionally, or implement deeper automation with AI solutions.
The Solution
2022
At the start, Mathema chose HelpCrunch as its main customer communication platform. Key reasons included:
- Support for multiple channels in a single interface — from messengers to social media
- Simple pricing model with no limits on the number of agents
- Automated chatbot workflows and trigger-based messages
- A knowledge base with AI-assisted editing and article adaptation
- Flexible API integration with their own CRM
This significantly reduced manual effort in handling requests, streamlined onboarding of new managers, and — most importantly — reduced the average first response time.
2025
In early 2025, Mathema became one of the first companies to test the beta version of AI Agents. They launched two specialized AI assistants, each trained on the knowledge base and tailored for different types of users:
- Assistant for parents and students — handles common questions (pricing, lesson prep, onboarding)
- Assistant for tutors — resolves typical operational queries from teachers
“We were pleasantly shocked by the result: 45% of chats are resolved entirely by the AI but without any support manager involvement. And that’s without a fully built-out knowledge base.”
Maryna Zaitseva, Head of Support & Operations at Mathema
What mattered most, however, was having the right tools to scale this automation.
- Knowledge Base — became a key onboarding tool for Mathema’s teachers, parents, and students.
- Pop-ups — the most effective way to quickly guide users to important product updates or features.
- API-powered Personalization — AI Agents greet users by name, know their roles, and tailor responses based on CRM data.
The Result
Since 2022, Mathema has gradually implemented HelpCrunch tools. After launching AI Agents in 2025, the company reached a fundamentally new level of customer service.
Key Achievements
- 45% of support requests are automatically resolved by AI Agents, allowing the team to focus on more complex, human-needed cases
- No negative impact on CSAT — user satisfaction remained consistently high post-AI rollout
- Support team size stayed the same, even though monthly chat volume grew 5x — from 6,000 in 2022 to nearly 30,000 by June 2025
- First Response Time dropped from 8–12 minutes to just 5 minutes
- The Knowledge Base became a highly used resource, with weekly views growing from 150 to 3,500+
| Metric | 2022 | 2025 |
|---|---|---|
| Number of chats/month | 6,000 | 30,000 |
| % of chats resolved without a manager | <5% | 45% |
| First response time (FRT) | 12 mins | 5 mins |
| Knowledge base views/week | 150 | 3,500 |
| Customer support team size | Unchanged |
HelpCrunch enabled Mathema to support 5x growth in support volume without growing the team, making scaling manageable and predictable.
“We’re handling 5x more support requests with the same team. That’s only possible thanks to deep automation.”
Maryna Zaitseva, Head of Support & Operations at Mathema
Future plans
Mathema plans to expand the use of AI Agents across all support channels and start using them for lead communication as well.
“We’re preparing to launch in another country — meaning another language and more new messengers. But we’re confident that with HelpCrunch, we can scale support processes quickly and cost-effectively.”
Maryna Zaitseva, Head of Support & Operations at Mathema
