Hi, I'm Pradnya!
I design AI systems that know their own limits, trust architectures where agents surface what they're confident about, stay silent when they're not, and learn from every human correction. 5+ years across enterprise AI, data-dense operational platforms, and design systems shipped to Intel, TSMC, Samsung, and Micron.
I believe the hardest AI design problem isn't confidence - it's uncertainty. My process is research-driven and prototype-heavy. I have rationalized 468K+ alarm signals into actionable intelligence and designed five confidence states that prevent AI anchoring bias.
Outside of work, I'm a national-level boxer, a dancer, and an eternal art lover, blending discipline and creativity in everything I do
-PhotoRoom%20(3)-PhotoRoom.jpg)
Selected Work
Fleet Intelligence: Designing Trust for a Three-Agent AI System
The hardest AI design problem isn't confidence, it's uncertainty. Five states calibrate when the AI speaks, when it stays silent, and how humans stay in control. Every correction makes it smarter
Impact
90%+
Agent Accuracy
AI-Driven Fleet Management: Predictive Analytics and Data Visualization for Optimizing Semiconductor Operations
Designing a Data-Centric Dashboard UI and Intuitive Query Builder for Uptime Analysis and Critical Parameter Inspection to Ensure Defect-Free Semiconductor Chips
Targeted Impact
+60%
Or higher accuracy in predictive maintenance
How others describe my work!
Peter Mckowen
Design Leader, Meta
"One of Pradnya's notable strengths was her dedication to running efficient and productive office hours. She created a welcoming and supportive environment where students felt comfortable asking questions and seeking clarification on course materials. Her responsiveness and attentiveness greatly contributed to the students' understanding of the subject matter. Additionally, Pradnya's commitment to providing weekly feedback on the course helped me continually improve the course throughout the semester.
Pradnya has a strong instinct for product design, and I strongly recommend her for any role that requires attention to detail, management of people, and a proactive approach to solving problems."