Case Study

DoiT empowers to architect an AI-ready infrastructure

Amazon SageMaker, Amazon Web Services


Acceleration in training time


Increase in productivity

Meet is a pioneering HR technology startup harnessing the power of artificial intelligence (AI) to transform how organizations manage their human resources. With its cutting-edge AI-powered platform, revolutionizes HR operations, delivering unparalleled efficiency, accuracy, and productivity throughout the talent management process.

By automating critical HR tasks such as recruitment, talent assessment, and training, enables businesses to unleash their full potential and cultivate high-performance teams. Leveraging advanced technologies like large language models and behavioral analysis algorithms, streamlines candidate screening, matching, and evaluation, accelerating hiring cycles and enhancing talent acquisition outcomes. stands out for its customizable and scalable solutions. The platform offers flexible configurations tailored to each organization’s unique requirements. Seamless integration with existing HR tools ensures a smooth user experience and optimal efficiency.

Trusted by leading corporations, staffing agencies, public institutions, and consulting firms, has quickly gained recognition for its transformative impact on HR operations. With as its partner, businesses unlock increased productivity, cost savings, and a competitive edge in attracting and retaining top talent.

The challenge faced the challenge of architecting an AI-ready infrastructure to future-proof its platform with cloud-native architecture and to optimize costs. To accomplish this, it needed in-house cloud expertise to manage the transition effectively. After evaluating different hyperscalers, chose AWS for its superior machine-learning capabilities, particularly SageMaker. However, it encountered significant obstacles when attempting to migrate its AI workloads.
Specifically, encountered difficulties running open-source large language models (LLMs) for natural language processing on SageMaker. The models failed during training due to version incompatibilities between SageMaker, PyTorch, and HuggingFace. This posed a risk of operational delays and hindered the deployment of the models.

To overcome these challenges, sought the support of DoiT, an AWS Premier Partner, for its expertise in cloud migration and machine learning. DoiT conducted a comprehensive assessment of’s architecture and identified the friction points within the system.

The solution

Working closely with’s engineers, DoiT conducted a meticulous assessment of the existing architecture, identifying improvement and potential bottlenecks. Based on this analysis, DoiT recommended a cloud-native approach using AWS services.

To ensure seamless integration, DoiT collaborated with’s team to determine the optimal combinations of Python, PyTorch, and HuggingFace versions compatible with AWS Deep Learning Containers. Through targeted training sessions, DoiT transferred this knowledge to empower’s team to independently manage and maintain its machine learning pipeline.

DoiT guided best practices for organizing the machine learning lifecycle, including development, storage, model deployment, and CI/CD automation. implemented a robust and scalable infrastructure for its AI initiatives by leveraging the ML Lens of the AWS Well-Architected framework.

Cost optimization was a key focus for DoiT. By enabling auto-scaling policies on SageMaker asynchronous inference endpoints, DoiT helped eliminate unnecessary resource consumption during idle periods while ensuring rapid scalability to meet spikes in demand.

Throughout the project, DoiT prioritized knowledge transfer and empowered’s team to own their AWS AI infrastructure. By doing so, gained the capability to operate and optimize its AI systems independently.

The following diagram illustrates the architecture implemented by for its resume categorization project, which effectively balances performance and cost considerations:

By partnering with DoiT International, successfully modernized its AI infrastructure, leveraging the power of AWS services and adopting industry best practices. The collaboration with DoiT has empowered to boost its productivity, streamline operations, and achieve its business objectives in the field of HR.

The result

Through the assistance of DoiT, successfully migrated its LLM workloads to AWS using SageMaker. This migration, coupled with’s AI-powered solutions, empowers HR teams to reduce their hiring time significantly by up to 60%.

By collaborating with DoiT, has positioned itself for scalability and cost optimization. Rather than simply transferring its existing infrastructure, DoiT enabled to embrace a cloud-forward architecture, swiftly implementing cutting-edge generative AI innovations and expanding its capabilities.

With a solid foundation built on AWS AI services, is committed to achieving market leadership through continuous improvement. DoiT will continue to guide, ensuring emerging best practices are adopted in responsible and trustworthy AI.

Verona Selimaj, International Development Lead,
“Through the assistance of Doit, we were able to successfully migrate our LLM workloads to AWS using SageMaker. Rather than simply transferring our existing infrastructure, DoiT enabled to embrace a cloud-forward architecture, swiftly implementing cutting-edge generative Al innovations and expanding our capabilities.
These innovations are helping us help our own customers, enabling global HR teams to reduce their hiring time by up to 60%.”

Learn more about how DoiT can help you

Latest case studies

Connect With Us