I turn the science of decision-making into financial products people trust.
Director-level behavioral science. I design the moments where money meets judgment — then run the experiments that prove the design moved real behavior.
Try a 15-second demonstration →Before the work, the thing itself. Two quick choices.
No right answer — pick the one you'd actually take.
Which would you take?
And this one?
Turning a missed payment into a resolved one
Removed the reassurance card that competed with the action
Less cognitive load — the card said "missed payments happen" which normalised avoidance
Added a caution cue
Saliency and urgency — naming the late-fee cost reframes the choice as a loss to avoid
Collapsed two competing buttons into one clear action
Fewer choices, faster resolution — choice overload suppressed action
Collections on missed payments were leaving money on the table. Standard dunning sequences — letters, robocalls, escalating fees — were generating regulatory risk without recovering debt.
Most missed payments aren't inability to pay — people forget, or the friction of calling in is too high. The emotional valence of 'collections' activates avoidance. Remove the threat, lower the effort, and most customers self-cure.
Secondary research, a behavioral & UX evaluation, and a three-variant click test (threat-based, neutral, friendly) on a panel of 400 users. The friendly variant was then A/B tested in production over 6 weeks.
Figures are rounded estimates. Mockups are illustrative — no real UI reproduced.
When congratulation backfired
Framed rewards as earned recognition for low earners
Motivational framing activates pride and status in lower-income segments
Removed congratulatory framing for high earners
For high earners, congratulation felt patronizing — insight from qualitative research
Rewards redemption rates were underperforming relative to enrollment. High-value customers were earning points they never spent — a liability on the balance sheet and a signal of low perceived value.
Redemption isn't a preference problem — it's a salience + timing problem. The value of rewards is highest at the moment of earning, not at the moment of redemption. Most redemption UX is buried behind menus and friction.
Contextual inquiry with 12 rewards cardholders, mapping emotional highs/lows in the rewards journey. Competitive teardown of 8 programs. Three low-fidelity concepts prototyped and tested with a 6-week live pilot of the winner.
Figures rounded. Mockups illustrative.
Matching people to places they actually want
Built preference-elicitation into the browsing experience
Stated preferences are unreliable; revealed preferences from behavior are more predictive
Used temporal discounting principles to rank urgency
People discount future options; surfacing high-match locations reduces present bias
Toyota's online configurator was losing customers mid-funnel. Analytics showed high drop-off between trim selection and lead submission — a 'paradox of choice' problem with 200+ configuration combinations.
Preference instability under complexity: when people can't hold all options in working memory simultaneously, they satisfice on salient features (color, price) and defer on others (tech packages, drivetrain). This creates regret risk and abandonment.
A 3-week diary study tracking preference stability across repeat visits, followed by a conjoint analysis to identify the minimum-viable configuration set. A progressive-disclosure prototype reduced options to a 4-question guided flow.
Figures accurate as of patent grant dates. Mockups illustrative — no real Toyota UI reproduced. No logos used.
Behavioral Research
Translating academic behavioral science into product hypotheses. From literature review through experiment design to result interpretation.
Experiment Design & Analysis
A/B and multivariate test design, statistical power, segmentation logic, and the discipline to report what didn't work alongside what did.
Team Building & Leadership
Hiring behavioral scientists and designers, building embedded team structures, and creating the organizational conditions where the work can ship.
I build the team and the practice, not just the research.
At Discover I hired behavioral scientists and product designers, owned the budget, and built embedded team structures that let the science reach the product surface. At Toyota R&D I developed IP that became five granted patents.
Kruti VekariaPhD · Georgetown University
I'm a behavioral scientist exploring how people make decisions — from the mundane to the extraordinary. My work bridges cognitive neuroscience, behavioral economics, and human-centered design to understand what drives generosity, health behavior, and financial well-being.
Contact
Get in touch
Book a call
[PLACEHOLDER: Cal.com booking calendar — available once CAL_LINK is configured]