.png)
Challenge
Jellysmack, a rapidly growing social media powerhouse with over 300 content creators generating 10 billion monthly views across 11 countries, faced a critical scaling challenge that threatened to bottleneck their exponential growth trajectory.
In-house limitations were holding them back. With 929 employees and a data science team that expanded from 7 scientists in 2016 to 35-40 by year-end, Jellysmack's existing in-house solution couldn't keep pace with their diverse AI use cases. The platform struggled to address various applications efficiently, from audience sentiment analysis and content classification to sophisticated video content editing across multiple distribution platforms.
Speed and scale became critical bottlenecks. As Andrea Colonna, Head of Data at Jellysmack, explained: "A critical issue at Jellysmack is to scale our capabilities quickly. Kili plays a big part as their solution enables our data scientists to be fully autonomous with building and managing training datasets and ship models into production faster."
Productivity and autonomy at stake. The company needed to ensure their data scientists could operate independently without constantly relying on engineering support or getting bogged down in project management tasks. The challenge wasn't just technical—it was about maintaining the agility and innovation speed that had driven Jellysmack's success in the competitive social media landscape.
Multiple AI use cases demanding attention. Jellysmack's AI requirements spanned two major domains: NLP applications including audience sentiment analysis and content classification, plus complex video processing for content editing across various social media platforms with different format requirements.
"We could have done it internally, but this task is very time-consuming, and we do not have a very efficient interface. We decided to partner with Kili to empower our data scientists."
Solutions
Jellysmack partnered with Kili Technology to implement a centralized, scalable training hub that would address all their AI use cases while maximizing data scientist autonomy and collaboration efficiency.
Centralized AI InfrastructureRather than maintaining separate solutions for different use cases, Jellysmack implemented Kili as their unified platform for both NLP and video ML applications. This strategic consolidation eliminated the inefficiencies of managing multiple tools and workflows across their diverse AI projects.
Intuitive and Advanced Annotation ToolsKili's platform provided Jellysmack's team with both user-friendly interfaces for quick onboarding and sophisticated annotation capabilities for complex projects. The dual approach ensured that new team members could contribute immediately while experienced data scientists had access to advanced features.
Streamlined Collaboration WorkflowsThe platform's collaboration features enabled seamless coordination between data scientists and labelers, eliminating previous bottlenecks in project management and data handoffs. This improvement was crucial for a team that had grown from 7 to nearly 40 professionals.
Autonomous Data Science OperationsBy providing data scientists with direct control over their training datasets and annotation processes, Kili eliminated dependencies on other teams and technical overhead that previously slowed down model development and deployment.
Efficient Onboarding and SupportAs Barthelemy Pavy, Data Scientist at Jellysmack, noted: "Kili makes it easy to collaborate. We can quickly onboard new project members, and take questions or provide feedback from and to our annotators to quickly solve issues or help them improve their skills."
Outcome
Jellysmack's implementation of Kili Technology delivered transformational results across their AI operations, enabling them to match their business growth with proportional increases in AI capability and efficiency.
Dramatic Speed Improvements The partnership enabled Jellysmack to ship models into production 2x to 10x times faster than their previous approach. This acceleration was critical for maintaining competitive advantage in the fast-paced social media industry where content trends and audience preferences shift rapidly.
Significant Efficiency Gains Data scientists now spend 50% less time on project management and collaboration overhead, allowing them to focus on core model development and optimization rather than administrative tasks.
Enhanced Model Performance Through continuous improvement workflows enabled by Kili's platform, Jellysmack achieved 4x-5x better model performance with ongoing optimization capabilities that ensure sustained improvements over time.
Reduced Operational Costs The streamlined approach delivered a 2x reduction in labeling costs while simultaneously improving quality and speed—a crucial efficiency gain for a company processing content at Jellysmack's scale.
Organizational Agility The platform enabled faster change management processes, helping Jellysmack move toward a more data-centric organization where AI insights drive strategic decisions across content creation and distribution.
Scalable Foundation for Growth Most importantly, Jellysmack now has the infrastructure to easily unlock new AI use cases as they emerge, ensuring their technical capabilities can scale alongside their business expansion across new markets and content formats.
"Kili had a real impact on our roadmap and the level of service we provide to our customers. Kili enabled us to improve our models' performance and scale our AI projects as fast as our business needs."
Future-Ready Operations With Kili's support for project management optimization and customer care, Jellysmack has established a robust foundation that supports both current operations across 11 countries and future expansion into new markets and content verticals.
Continuous Innovation Pipeline The platform's intuitive interface and advanced capabilities ensure that as Jellysmack continues to grow their team of data scientists, new members can quickly contribute to ongoing projects while experienced team members can push the boundaries of what's possible in social media AI applications.
.png)
