Welcome.AIWelcome.AI
    Skip to content
    Speech Recognition

    Tarteel Achieves 56% Cost Savings and 22% Latency Reduction with CoreWeave Migration

    Tarteel AI's transition to CoreWeave marks a significant enhancement in cloud performance, achieving a 22% reduction in latency while saving nearly 56% in costs, setting a new standard for niche cloud solutions.

    wf.coreweave.com•October 2, 2026•2 min read

    Key Facts

    • Tarteel's 56% cost reduction post-migration highlights financial efficiency in cloud strategy.
    • 22% latency improvement indicates competitive edge in user experience over legacy providers.
    • Access to NVIDIA GPUs reveals CoreWeave's strategic advantage in AI infrastructure flexibility.
    • Zeet's seamless deployment minimizes downtime, showcasing operational efficiency in cloud transitions.
    • Tarteel's growth potential in NLP and TTS indicates strategic market expansion opportunities ahead.

    Summary

    Summary

    Tarteel AI, an AI-powered Quran study app, migrated its cloud infrastructure from AWS to CoreWeave with assistance from Zeet. This transition resulted in a 22% improvement in latency and approximately 56% cost reduction, enhancing the app's performance and affordability.

    Background

    Tarteel is an AI-driven application designed to assist over 2 billion Muslims worldwide in their Quran study. Prior to the migration, Tarteel relied on AWS for its cloud services but faced limitations in GPU selection, observability, and overall performance for its production workloads.

    Challenge

    The primary challenge for Tarteel was the inadequacy of legacy cloud offerings, which lacked observability for workloads and provided an unoptimized solution for inference and production tasks. This hindered Tarteel's ability to scale and improve its service.

    Solution

    Tarteel migrated its entire cloud infrastructure to CoreWeave, leveraging Zeet for a seamless transition. The deployment included access to a wide range of NVIDIA GPUs and tools like NVIDIA Riva and NeMo, which facilitated the training and deployment of state-of-the-art ASR and TTS models. The integration with Zeet simplified the setup process, allowing Tarteel to focus on product development.

    Results

    Post-migration, Tarteel achieved significant performance improvements, including:

    • An average of ~1,600 requests per minute across over 40 NVIDIA GPUs (A4000/5000s).
    • A 22% reduction in latency, with a median request latency of 423.5 ms.
    • Approximately 56% reduction in operational costs.

    Key Insights

    Businesses considering cloud migration should evaluate the flexibility and performance of potential providers. The integration of advanced tools and support can significantly enhance operational efficiency and reduce costs. Additionally, leveraging a partner for deployment can minimize downtime and technical challenges.

    Customer Testimonial

    “Zeet made the switch to CoreWeave a breeze. We had minimal downtime and did not need any K8s or CW expertise, we just clicked a few buttons and we were live.” – Anas Abou Allaban, Co-Founder and CEO, Tarteel

    Entities Mentioned

    Companies

    Tarteel
    Zeet
    CoreWeave

    Products

    Tarteel AI
    NVIDIA Riva
    NVIDIA NeMo

    Technologies

    NVIDIA GPUs
    Docker
    Kubernetes
    Grafana
    Prometheus
    Triton

    People

    Anas Abou Allaban

    Key Concepts

    cloud migration
    AI-powered applications
    cost reduction
    latency improvement
    GPU selection
    observability
    autoscaling
    NLP and TTS

    Definitions

    cloud migration
    The process of moving data, applications, and other business elements from on-premises infrastructure to a cloud environment.
    observability
    The ability to measure and understand the internal state of a system based on the data it produces.
    NLP
    Natural Language Processing, a field of AI that focuses on the interaction between computers and humans through natural language.
    TTS
    Text-to-Speech, a technology that converts written text into spoken words.
    autoscaling
    The ability of a system to automatically adjust its resources based on the current demand.

    Use Cases

    • →real-time Quran recitation correction
    • →semantic search for Quranic texts
    • →voice-guided Quran study
    • →NLP for Q&A
    • →spaced repetition for chapter recitation
    • →digitizing Islamic/Arabic literature

    Frequently Asked Questions

    What benefits did Tarteel gain from migrating to CoreWeave?

    Tarteel experienced a 22% improvement in latency and approximately 56% cost reduction after migrating to CoreWeave. The platform also provided better GPU options and observability features.

    How did Zeet facilitate Tarteel's migration?

    Zeet simplified the migration process by quickly resolving issues related to load balancing and DNS setup. This allowed Tarteel to complete the migration in just 1-2 days with minimal downtime.

    What technologies does Tarteel utilize for its AI functionalities?

    Tarteel uses NVIDIA's Riva and NeMo toolkits to enhance its AI capabilities, allowing for improved performance in speech recognition and text-to-speech applications.

    What is the primary function of the Tarteel app?

    The Tarteel app is designed to help users practice and improve their Quran recitation through interactive and voice-guided features, providing real-time feedback on their performance.

    What future plans does Tarteel have regarding AI?

    Tarteel aims to leverage AI for creating personalized experiences related to Quran study, including semantic search, Q&A functionalities, and digitizing important Islamic texts.

    Where AI Leaders Stay Informed

    The latest AI intelligence, case studies, and research — delivered to your inbox every week.

    Free to read. Unsubscribe anytime.