AI-Powered Recruitment Transformation: PandoLogic's Journey with Automated MLOps Pipelines

PandoLogic is an AI-based programmatic job advertising platform that intelligently automates and optimizes job advertising spend. Their platform, pandoIQ, provides an end-to-end job advertising solution that delivers a significant increase in job ad performance without any wasteful spending. PandoIQ’s AI-enabled algorithms use over 200+ job attributes and more than 200 billion historical job performance data points to predict the optimal job advertising campaign. The models are continuously being retrained automatically to production. PandoLogic serves customers like Fedex, Dominos, Postmates and other Fortune 500 organizations.

PandoLogic, an AI-based programmatic job advertising platform, was facing challenges in operationalizing machine learning (ML) on premises due to lack of DevOps or infrastructure. Despite having a lean team of data scientists, their productivity was constantly interrupted with DevOps tasks and infrastructure challenges. They lacked the resources and infrastructure to operationalize their impressive models to achieve real business results. They were limited to on-premise deployment which caused technical challenges and DevOps overhead. They required dynamic Spark Clusters to handle terabytes of data, which wasted weeks of set up time, and caused major overhead costs to maintain. The PandoLogic team wanted a way to train and deploy on premise and in multiple clouds, without being locked into a single cloud. They needed an easy way to leverage open source tools and the compute resources they already had.

PandoLogic adopted cnvrg.io, an enterprise-level ML infrastructure platform, to overcome their challenges. cnvrg.io provided one-click integration with AWS data lakes, Kubernetes deployments, Spark Clusters, and cloud compute all with little overhead. It delivered a resource management system that can leverage on-prem resources and automatically bursts to the cloud and terminates with one click. PandoLogic leveraged cnvrg.io’s Flow UI tool to build custom ML pipelines that allows each task to run on a different cloud simultaneously. They were able to take their Jupyter Notebooks to process the data, save metadata and analyze the flow, while using Python to run experiments, and serve their models with Kubernetes, AWS and on-premise servers. Using the AI library, PandoLogic built custom drag and drop pipelines pulled from their GitHub repo, and used their proprietary algorithms as well as XGBoost, LightGBM, Ensemble Methods and other open source ML and DL frameworks like PyTorch with the cnvrg.io platform.

• With cnvrg.io, PandoLogic was able to quickly adopt an entire enterprise-level ML infrastructure platform from zero to production. This would have otherwise taken a team of specialized ML engineers, and months of DevOps time to build. cnvrg.io’s MLOps solution has accelerated workflows drastically, allowing spinning up a powerful AI environment in one click without manual DevOps. PandoLogic can now version control data, experiments and monitor models in a single place, something their previous version control system was not comprehensive enough to handle. This has allowed PandoLogic to focus on building and delivering models rather than building their own infrastructure, thus adding real business value.

• not mentioned