Pramil Panjawani

Hiring Analytics Dashboard (Ontology + GraphDB)

• Spearheaded the development of an ontology-driven hiring analytics dashboard that unified disparate recruitment data sources to analyze pipeline performance, recruiter effectiveness, and time-to-hire.
• Designed graph-based data models in Neo4j, leveraging Apache Iceberg for scalable storage and Trino for querying multi-source data.
• Enabled visibility into hiring metrics segmented by teams, departments, and roles, facilitating bottleneck diagnosis and strategic decision-making for enterprise clients.

User Segmentation & Personalized Recommendations

• Developed a user segmentation framework to personalize job and candidate recommendations based on recruiter behavior and activity history.
• Applied clustering techniques and engagement signals to group users by interaction patterns, using Pandas and scikit-learn for modeling.
• Deployed segmentation-powered recommendation logic that increased interaction rates on surfaced content by 8–18% across key user cohorts.

Career Path & Role Progression Generator

• Built a system to generate career path and progression recommendations within organizations using historical transitions, skill clusters, and job metadata.
• Used Pandas to extract progression patterns and applied LLMs for semantic validation of recommended paths, ensuring realism and domain alignment.
• Supported use cases like internal mobility and upskilling by offering multiple validated next-step suggestions for each role.

Candidate Filtering Pipelines (Mongo + Kafka)

• Developed candidate filtering pipelines by ingesting data from MongoDB, processing it with Pandas, and publishing filtered results via Kafka.
• Implemented filters based on loyalty score, alumni status, experience similarity, and other client-specific rules.
• Built a configurable scheduler to automate data updates per client logic (e.g., blocked companies/locations), improving recruiter efficiency and relevance of candidate pools.