- Growing AI apps with influencers: investor Paul Shaburov on building GlamAI, ChatChat & going viral
- Published 25/04/2025
- 6 min Read
- Written by Julia Shandrokha
Table of contents
We’re excited to feature Paul Shaburov — founder of GlamAI and ChatChat, and partner at MLVentures, experienced mentor and active investor — on our blog.
Paul brings deep experience in building and scaling AI-powered consumer apps, with a track record that includes exits to Snap, Opera, and Yandex. In this interview, he shares what it takes to launch products like GlamAI, how his team experiments with influencer marketing, and why rapid iteration and smart positioning matter more than ever.
If you’re working on AI, mobile, or creator-driven products, you’ll find a lot of practical insight here.
Paul Shaburov’s experience & expertise
Can you briefly introduce yourself and your background in tech/AI? What led you to start MLVentures and later launch GlamAI and ChatChat?
I have a background in machine learning and have been involved in building and scaling several AI-driven products. In the past, I played a key role in the acquisitions of AI Factory and Looksery, which were acquired by Snap for $166 million and $150 million, respectively.
I started MLVentures as a continuation of our previous family fund. We partnered with strong technical founders, which later resulted in the launch of GlamAI and ChatChat.
You’ve had successful exits in the past, please tell us a bit about them. How has that experience shaped your approach to launching and scaling new products today?
- Looksery — The company developed a real-time facial modification app for mobile platforms. In September 2015, it was acquired by Snap Inc. for $150 million, marking the largest startup acquisition in Ukraine’s history. Snapchat used Looksery’s technology to launch the Lenses feature.
- AI Factory — A computer vision startup acquired by Snap for $166 million in 2019 to power Snapchat’s animated selfie-based Cameos feature.
- SPB Software — One of the largest mobile software developers, with products like SPB TV, Pocket Plus, and Mobile Shell. It was acquired by Yandex in 2011 for €38 million.
- Handster Inc. — A major developer of mobile app store platforms, acquired by Opera in 2011.
These past exits taught me to prioritize fast iteration, user-centric design, and strategic alignment — building products that scale quickly and attract meaningful partnerships or acquisitions.
What’s the core idea behind MLVentures? How do you choose which products to build or invest in?
The core idea behind MLVentures is to identify and seize opportunities in the rapidly growing AI app industry by applying a strong, data-driven business approach to product development. We focus on creating AI-powered applications that deliver unique and personalized user experiences based on consumer preferences.
We choose what to build or invest in by selecting innovative ideas with strong potential and supporting them from concept to market expansion. Our focus is on backing technically skilled founders — such as ML experts, ACM ICPC finalists, and Kaggle medalists — who are using cutting-edge AI technologies to build the next generation of AI-driven solutions. While we remain flexible, our priority is always on high-impact, scalable products in the AI space.
GlamAI – Product & growth
What inspired the creation of GlamAI, and how did it evolve into a broader fashion/creative platform?
Many product iterations and pivots. Eventually, we found a model that is both sustainable and profitable.
What do you think made GlamAI go viral so quickly — 1M+ downloads and profitability in just a few months?
We added cool styles tailored to a B2B segment, which resulted in very low CPMs in our marketing campaigns. That allowed us to scale to over 1M downloads.
You’ve built for both B2C and B2B — how different are those audiences, and how do you approach user acquisition for each?
Currently, 100% of our revenue is coming from B2C. We are only starting B2B now. For us, B2B is something that is still yet to be conquered.
What technologies power GlamAI behind the scenes? How is your approach to try-on different from traditional AR or 3D solutions?
We’ve developed many of our own foundational models. Our try on is fully a 2D solution.
ChatChat – Concept & use cases
What’s the idea behind ChatChat, and how does it stand out from other AI chat platforms?
I think the main idea behind ChatChat is combining live conversation with game-like mechanics. This lets users interact without getting stuck in templated dialogues, while still keeping the flow and initiative. Copilots are a great feature, and we also have a lot of AI-generated content — images, voice messages, and more.
What sets ChatChat apart from other AI chat platforms is its unique blend of features:
- AI-powered matchmaking connects you with virtual companions who match your interests — whether you’re looking for a friend or a romantic partner.
- Emotional expression with lifelike actions like hugs and kisses creates a warm, realistic feel.
- Custom character creation lets you design unique AI personas that reflect your imagination.
- Dynamic AI-generated scenarios keep conversations natural, engaging, and full of surprises.
- Personalized goal progression tailors your journey, making each interaction rewarding.
- Immersive roleplay turns chats into adventures — fun, emotional, and endlessly creative.
What are some of the most interesting or unexpected ways people are using ChatChat (for work, creativity, etc.)?
Probably the most unexpected use I’ve seen is our users creating custom bots of people who have passed away, so they can keep talking to them. Some also create customs based on their pets. There’s not a huge demand for AI helpers, since our ads mainly focus on the social aspect of the platform.
ChatChat feels privacy-first and community-driven — was that intentional from the start, and how do you balance openness with responsible AI use?
I’m not entirely sure what you mean by “responsible AI use,” but we do have a strong moderation system in place — for both text and image content. It filters not only AI-generated material but also user-created content.
Influencer marketing – strategy & insights
What role has influencer marketing played in promoting GlamAI? Did it contribute to its viral growth?
Influencer marketing helped us achieve a viral effect in Latin America and several European countries with positive ROI. Creator content also performs well in Spark Ads.
How do you choose which influencers to work with — is it mostly beauty/fashion creators, or do you experiment with other niches?
We work with micro-influencers in our target geos who have high engagement rates and low cost per video, without focusing on specific content niches.
What platforms (e.g. TikTok, Instagram, YouTube) have been the most effective for your influencer campaigns? Why?
TikTok. We create viral content that shows photo and video results without revealing the app itself. This approach shows better results than traditional campaigns.
Do you focus more on organic/user-generated content or paid collaborations with influencers?
We focus on organic and user-generated content but also pay creators for collaborations.
Could you share any examples or results from successful influencer campaigns — e.g. uplift in downloads, MAU growth, viral moments?
Our best monthly result: 40M views, CPM = $0.20, estimated CPI = $0.04–$0.07, CAC between $1–2.50.
Have you experimented with using influencers to promote ChatChat as well, or is that product marketed differently?
We’ve used influencers to promote ChatChat, but the results weren’t very meaningful. Now we use standard performance marketing instead.
Strategy, challenges & lessons
What have been the biggest challenges in scaling apps like GlamAI and ChatChat through influencer-led growth?
The biggest challenge is measuring ROAS, as the traffic coming to the apps is very difficult to segment.
How do you measure ROI or success when it comes to influencer marketing campaigns?
We don’t use tracking links, so we measure success by monitoring the growth in organic installs.
What advice would you give to other app founders and developers who are considering working with influencers to grow their apps?
Post on TikTok, test a variety of creative concepts and scale what works well.
What’s next for you and the MLVentures portfolio — any upcoming projects or new areas you’re exploring?
We would like to explore more speech-based mobile products — for example, building the next Suno. Suno AI is a generative music platform that enables users to create full songs — including vocals, lyrics, and instrumentation — by simply entering a text prompt.
Wrapping up
Big thanks to Paul for taking the time to share his experience — from early pivots to viral campaigns and influencer strategies that actually move the needle.
For founders and marketers alike, his answers offer a real-world look at growing AI apps beyond hype. Whether you’re exploring GenAI, launching a mobile product, or scaling with creators, there’s something here to take away.
Stay tuned — we’ll keep bringing more behind-the-scenes stories from people shaping the future of tech and content.
* * *
Want to be the next hero in our series?
If you’re building something exciting and have a story worth sharing, we’d love to hear from you. Drop us a message — and let’s feature your success next: [email protected]
TAGS :
