Deskcrit App - Critique that argues back.
Read the story
Ravi Palwe · Senior Manager, Experience Design
I design thetrust layer betweenpeople andintelligent systems.
20 years designing for banks, insurers and asset managers. Since 2024, designing the AI systems they're finally letting near the money. These industries adopt this technology last, and they have good reasons for it.
Greater Richmond, VA · Senior UX · Full-time, contract, remote, or hybrid
- 20yrsFinancial services UX
- 2AI productsDesigned, built, and live
- 5papersPeer-reviewed
- 35%fasterTask completion, AI app
- 25%liftEngagement, AI app
Writing
Papers & Articles
- PaperComprehensive PublicationsDesigning the Invisible: UX Strategies for Agentive AI in 2025 and Beyond (opens in a new tab)Design patterns for AI that acts on the user's behalf: trust calibration, explainability thresholds, and the right to override.2025
- PaperArtificial Intelligence Journal | Computing JournalRedefining Usability in the Age of Generative AI: Towards a New Evaluation Paradigm (opens in a new tab)Traditional usability heuristics break when the interface generates novel output. Updated evaluation criteria for systems where outcomes are non-deterministic.2025
- PaperInternational Journal of Computing and Artificial IntelligenceOnboarding for AI Features: Reducing Friction at the First Use (opens in a new tab)How first-use friction creates lasting skepticism, and the design patterns that reduce abandonment during onboarding.2025
- PaperIJISAEAnalyzing the Core Principles and Methodologies of Ethnographic UX Research in Large Corporate Contexts (opens in a new tab)Conducting ethnographic research inside regulated enterprises, where observation access is restricted.2023
- PaperIJSRThe Role of User Experience Design in Advancing Digital Sustainability (opens in a new tab)Introduces SOLAR, an open-source framework for putting environmental cost into component-level design decisions.2025
About me
Hi, I'm Ravi
I've spent 20 years getting people to trust screens with their money.
I lead AI/UX transformation for financial services at Capgemini, designing for one of the world's largest wealth managers, two of the biggest US banks, and a Fortune 500 specialty insurer. Regulated work is slower. There is always someone who has to sign off. That constraint made me better at this: you cannot ship your way past a bad decision, so you have to argue it properly the first time.
Most of my time now goes to agentic AI. These systems don't chat or recommend. They go and do the thing. Once software can rebalance a portfolio or settle a claim without asking first, most of the design work moves: what the system is allowed to assert, what it has to show you before it acts, and how a person takes a decision back.
I still ship. The critique agent above went from brief to production in a week, built by me. Right now I'm building Throughline, a research synthesis tool where every AI-generated insight traces back to its verbatim source. Outside the work: medium-format film, 6×6 or 6×7, and I don't crop. Ever.
"I don't believe in cropping in photography, and I don't believe in cutting corners in product design."
My design philosophy