DoiT empowers Sweeeft.ai to architect an AI-ready infrastructure
Meet Sweeeft.ai
Sweeeft.ai 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, Sweeeft.ai 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, Sweeeft.ai enables businesses to unleash their full potential and cultivate high-performance teams. Leveraging advanced technologies like large language models and behavioral analysis algorithms, Sweeeft.ai streamlines candidate screening, matching, and evaluation, accelerating hiring cycles and enhancing talent acquisition outcomes.
Sweeeft.ai 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, Sweeeft.ai has quickly gained recognition for its transformative impact on HR operations. With Sweeeft.ai as its partner, businesses unlock increased productivity, cost savings, and a competitive edge in attracting and retaining top talent.
The Challenge
Sweeeft.ai faced the challenge of architecting an AI-ready, cloud-native foundation that future-proofs its platform and optimizes costs. To accomplish this, it lacked sufficient in-house cloud expertise to manage the transition effectively. After evaluating different hyperscalers, Sweeeft.ai chose AWS for its ML stack—particularly Amazon SageMaker. However, it encountered significant obstacles migrating AI workloads to Amazon SageMaker. Specifically, Sweeeft.ai 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 Deep Learning Containers (DLCs), PyTorch, and Hugging Face Transformers. This threatened timelines and blocked model deployment.
To overcome these challenges, Sweeeft.ai sought the support of DoiT, an AWS Premier Partner, for its expertise in cloud migration and machine learning. DoiT conducted a comprehensive assessment of Sweeeft.ai’s architecture and pinpointed integration friction and version drift.
The Solution
Working closely with Sweeeft.ai’s engineers, DoiT conducted a meticulous assessment of the existing architecture, identifying areas for improvement and potential bottlenecks. Based on this analysis, DoiT recommended a cloud-native AWS approach aligned to the target workloads.
To ensure seamless integration, DoiT collaborated with Sweeeft.ai’s team to determine the optimal combinations of Python, PyTorch, and HuggingFace versions compatible with AWS Deep Learning Containers. Through targeted training, DoiT enabled Sweeeft.ai’s team to own and operate its ML pipeline independently.
DoiT guided best practices for organizing the machine learning lifecycle, including development, storage, model deployment, and CI/CD automation. Sweeeft.ai 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 Sweeeft.ai eliminate unnecessary resource consumption during idle periods while ensuring rapid scalability to meet spikes in demand.
Throughout the project, DoiT prioritized knowledge transfer, empowering Sweeeft.ai to own its AWS AI infrastructure. By doing so, Sweeeft.ai gained the capability to operate and optimize its AI systems independently.
The following diagram illustrates the architecture implemented by Sweeeft.ai for its resume categorization project, which effectively balances performance and cost considerations:
By partnering with DoiT, Sweeeft.ai modernized its AI infrastructure, leveraging AWS services and adopting industry best practices.The collaboration with DoiT has empowered Sweeeft.ai to boost its productivity, streamline operations, and achieve its business objectives in the field of HR.
The Results
With DoiT’s assistance, Sweeeft.ai migrated LLM workloads to Amazon SageMaker on AWS. This migration, coupled with Sweeeft.ai’s AI-powered solutions, helps HR teams cut hiring time by up to 60%.
By collaborating with DoiT, Sweeeft.ai is now positioned for scale and cost efficiency. Rather than a lift-and-shift, DoiT enabled a cloud-forward architecture to accelerate GenAI adoption and expand capabilities.
With a solid foundation on AWS AI services, Sweeeft.ai is pursuing market leadership through continuous improvement. DoiT will continue advising Sweeeft.ai on adopting emerging best practices in responsible, trustworthy AI—from versioning strategies to scalable, cost-aware deployments.
These innovations are helping us help our own customers, enabling global HR teams to reduce their hiring time by up to 60%.”