AI data pipelines
Delivery lead for teams fine-tuning AI models on proprietary customer data.
Program delivery
~30% faster
3model families: Gemini, ChatGPT, Claude
~30%less data processing time
Dailyexecution syncs
C-suiteto data scientist
The problem
- Teams were building and fine-tuning AI models on customer datasets.
- Research needs were ambiguous and priorities kept moving.
What I did
- Served as Scrum Master across Scrum and Kanban workstreams.
- Turned research needs into backlogs, plans and measurable KPIs.
- Found the bottlenecks in the data and review pipelines and re-sequenced the work.
- Ran daily execution syncs to surface issues early.
What changed
- Data processing time down about 30%.
- More predictable delivery.
- Clear reporting to senior stakeholders.