Personalizing talent acquisition to improve hiring outcomes
104 Corporation· TaiwanDelta Lake · MLflow · Databricks Lakeflow Jobs
104 Corporation, Taiwan's leading online recruitment platform, moved data workflows to the Databricks Data + AI Platform and deployed MLflow to train machine learning models that summarize job postings and power personalized, tailored recruitment recommendations for job seekers and employers. The company saw a 25% reduction in ETL processing time and 9x faster data processing for more personalized job-seeking experiences.
Overview
104 Corporation, Taiwan's leading online recruitment platform, moved data workflows to the Databricks Data + AI Platform and deployed MLflow to train machine learning models that summarize job postings and power personalized, tailored recruitment recommendations for job seekers and employers. The company saw a 25% reduction in ETL processing time and 9x faster data processing for more personalized job-seeking experiences.
This entry has 14 published fields tied to exact passages in an immutable source capture.
Inspect the highlighted sourceThe challenge
104 Corporation struggled to scale pipeline costs effectively and spent significant resources on maintenance operations, slowing their ability to be truly data-driven, hampered by specialty ETL tooling like EMR and a legacy data warehouse in SAP Sybase that were disparate and complex to scale.
The solution
104 Corporation moved data workflows onto the Databricks Data + AI Platform and Delta Lake, building high-performance ETL pipelines that support training new ML models to better summarize company and job postings. It separately deployed MLflow, which brought data scientists, engineers and developers into one unified ecosystem for collaboration and knowledge-sharing.
Reported business value
104 Corporation has seen an average 25% reduction in the amount of time needed to process ETL workloads for both BI dashboarding and ML model training, with runtime for high-volume analytics decreased to around a minute.
Sources
Open any source and check the claim yourself — that is the point of the register.
This record was researched and written with AI assistance, and its claims were checked against the sources above. (EU AI Act art. 50 transparency notice.)
Other human resources entries in the register.
Workday Accelerates Generative AI / ML Product Development Using Amazon SageMaker
Workday uses Amazon SageMaker to let engineering teams build, train and deploy ML models, including LLMs, across AWS Regions to meet global customers' data-residency requirements. Using Amazon SageMaker Studio, Jumpstart and Ground Truth Plus, Workday piloted a closed-book ML application analyzing job descriptions, invoices and contracts, improving ML inference latency by a factor of five. Workday also received early access to Amazon Bedrock for generative AI prototyping. Head of Workday AI Shane Luke said the company has scaled from a thousand inference requests to tens of millions coming in daily.
WorkdayAmazon SageMaker · Amazon SageMaker Studio · Amazon SageMaker Jumpstart +2Checkr: Streamlining Background Check Classification with Fine-tuned Small Language Models
Checkr automated adjudication of background checks using a tuned logistic regression model for 98% of cases, but the remaining 2% required classifying noisy records into 230 categories. After testing GPT-4 (80-82% accuracy on hard cases, ~15s latency, ~$12k cost) and GPT-4 with RAG, Checkr fine-tuned Llama-2-7b (85% accuracy) and then, in production on Predibase, Llama-3-8b-instruct, achieving 90% accuracy on the hardest 2% of cases with 0.15-second response times (30x faster than GPT-4) and a 5x cost reduction, using LoRA/LoRAX for parameter-efficient fine-tuning and multi-adapter serving.
CheckrLlama-2-7b · Llama-3-8b-instruct · LoRA +3Was this helpful?
Your feedback helps us improve our use case database