Skip to content
All work

Multi-modal data → insight

Joblogic

Customer Data Analysis Platform

Unstructured, multi-modal comms turned into clustered insight at scale.

Role
AI Engineer — design & build
Year
2025
Access
Private — walkthrough on request
  • Embeddings
  • Clustering
  • Transcription
  • LLM analysis

What I built

A Databricks pipeline that processes customer communication archives through transcription, embeddings, clustering, and LLM analysis. It groups related conversations and turns representative samples into structured insight for automation planning.

My role

I designed and built the data pipeline end to end as an AI Engineer at Joblogic.

Shown at a conceptual level — no proprietary architecture or detail.

Next case study

CallCover
Work with me