Biography will be updated soon.

Jebaraj Vasudevan is a seasoned Data Scientist with a track record of impactful work in Applied Data Science. In his current role as Senior Manager, Data Science at Visa Inc., Jebaraj applies advanced analytics to uncover business-critical insights, directly contributing to significant revenue growth for the organization. His prior experience at Bosch involved leveraging statistical and data science methods to enhance product development, resulting in tangible improvements for consumers. Jebaraj’s academic background includes a Masters in Data Science from the University of Texas at Austin and a Masters in Mechanical Engineering with a minor in Computational Engineering from Purdue University. Beyond professional achievements, Jebaraj is also an active member of the tech community, serving as an IEEE senior member, peer-reviewing papers for leading conferences such as NeurIPS and SciPy. Reference: Two-Step Validation for Observational FinTech Causal Estimates: Retroactive/Pseudo-Event Falsification and Synthetic-Control Alignment via Covariate Matching and Temporal Trajectory Methods Observational causal estimates in financial technology (FinTech) require explicit validation beyond identification to ensure credible inference. This review formalizes a two-step validation paradigm in which an initial causal estimate is followed by a falsification-driven validation stage. We synthesize two validation mechanisms: (i) retroactive and pseudo-event validation based on temporal falsification, and (ii) synthetic-control validation enabled by counterfactual alignment through covariate matching and temporal trajectory matching. When suitable controls exist, KD-tree nearest-neighbor search coupled with entropy balancing enforces pre-treatment covariate similarity and supports robust weighted difference-in-differences estimation. When direct controls are limited, Dynamic Time Warping aligns temporal patterns while Bayesian Structural Time Series constructs latent counterfactual paths through state-space modeling. Together, these methods address imbalance, temporal misalignment, and structural uncertainty, providing a technically grounded validation framework for causal inference on observational FinTech datasets.