Talkpal Scaled on AWS with a More Cost-Efficient Setup

Back

The Client

Talkpal is an AI-powered language learning platform that helps users practice more than 80 languages through immersive, interactive experiences.

Learners can improve by writing or speaking on a wide range of topics while receiving responses in a realistic voice, making practice feel natural and engaging. With features like Chat, Roleplays, Characters, Debates, and multiple learning modes, Talkpal creates a highly personalized learning experience across web, iOS, and Android.

Industry:Education
Company Size:11-50
Country:United States

“Cloudvisor has been a valuable partner in supporting Talkpal’s AWS and DevOps infrastructure as we scale globally. Their team has helped us build a more reliable, efficient, and scalable cloud environment, allowing us to focus on delivering the best possible AI-powered language learning experience to our users.”

Mikheil Andguladze

DevOps Team Lead

Challenges

Unclear savings potential – The team wanted to understand how much they could realistically save using Reserved Instances and Savings Plans, but lacked a clear breakdown based on their actual usage.
Understanding Database Savings Plans – With the introduction of Database Savings Plans, it wasn’t obvious which configurations or instance types would allow them to fully benefit from the new model.
Need for deeper analysis beyond COR – While the initial Cost Optimization Review provided direction, it didn’t fully answer all practical questions around implementation and commitment levels.
Infrastructure alignment for savings – To take advantage of Savings Plans, certain adjustments in instance types and database configurations were required, which needed careful evaluation.
Balancing cost optimization with growth – As Talkpal continued to scale, the goal wasn’t just to reduce costs, but to ensure spending was efficient and aligned with long-term usage.

Before working with Cloudvisor, Talkpal had a solid AWS setup in place, but needed deeper visibility and guidance to confidently implement Savings Plans and optimize their infrastructure without limiting future growth.

Solutions

Cloudvisor started with a cost review and then worked closely with the team to define practical ways to apply Savings Plans effectively based on their real usage.

Cost review and opportunity mapping

An initial assessment was carried out to highlight where savings could be captured across key services, with a focus on compute and database workloads.
Detailed savings simulations

Custom calculations were prepared to compare different commitment options, helping Talkpal understand how various scenarios would affect their costs over time.
Planning compute commitments

Guidance was provided on selecting the right Savings Plans coverage for compute resources, ensuring efficiency without overcommitting.
Enabling database-level savings

Cloudvisor advised on adjusting database configurations so they could benefit from the newer Database Savings Plans model.
Aligning infrastructure with Savings Plans requirements

Recommendations focused on adjusting existing resources to meet Savings Plan requirements, ensuring the setup supports cost optimization without disrupting operations.
Customised EKS Node Configuration

EKS node AMIs were customised to pre-load Docker images and ML models, simplifying automation and reducing initialization time for AI workloads.
Ongoing collaboration and support

After sharing the analysis, Cloudvisor continued working with the team to answer follow-up questions and support the implementation of the selected approach.

AWS Services Used

Amazon EC2
Amazon RDS
Amazon EKS
AWS Data Transfer

Results

Following the analysis and implementation of the recommended changes, Talkpal was able to make more informed decisions around their AWS spending and infrastructure setup. While overall AWS spend increased alongside growth, their setup became more cost-efficient and better aligned with their usage, supporting both current needs and future scaling.