Praatyush Gupta

-Ran user research across Adobe-sponsored campus events to assess adoption of new AI tools (Firefly), identifying onboarding and feature-discovery gaps driven by low student awareness. -Re-prioritized event focus and product messaging around Firefly and Adobe Creative Cloud integration, contributing to a 15% increase in student engagement over the semester.

- Facilitated learning experiences by leading 50-minute lectures for 30 students twice weekly, conveying concepts involving computer systems, object oriented programming, data structures, and algorithms - Conducted comprehensive office hours, mentoring up to 900 students with debugging, clarifying course material, teaching design patterns, unit testing, code optimization, and providing guidance on programming challenges

-Led end-to-end development of an LLM-driven competitor intelligence platform for a $4B pharma client, transforming unstructured financial and clinical reports into product-level oncology insights. -Drove stakeholder alignment across senior leadership and healthcare experts, translating AI capabilities into executive-grade insights that reduced manual competitive analysis time by ~80%.

- Led 0 → 1 launch of a swipe-based B2B app, leveraging an intuitive engagement loop to connect jewelry wholesalers with retailers, increasing brand engagement by 17% and revenue by 8%. - Established key success metrics and conducted A/B testing on swipe vs. grid-based browsing, boosting product discovery by 20% and improving conversion rates. - Led cross-functional collaboration with 2 engineers, 1 designer, and stakeholders, iterating on user insights to enhance UX, increasing retailer retention by 12%.

- Architected a Python SDK & logging framework in Databricks for secure AWS Redshift access using secrets, cutting credential incidents by 18% and troubleshooting time by 24% - Scaled across 6 global workspaces, standardizing partitioning and sort/distribution keys on multi-TB Redshift tables to slash query latency by 10% - Developed Spark ML models for dynamic Redshift autoscaling and anomaly detection, forecasting workload patterns to achieve a projected 3% cost reduction while ensuring high availability - Deployed and integrated the library across AARP, where it was adopted by over 50 employees, leading to a 30% increase in adherence to best practices for secure and efficient data management.

- Spearheaded development of a course management portal for Y-KNOT using React.js, TypeScript, CSS Modules, and Firebase, boosting user engagement by 40% and cutting administrative workload by 30% - Leveraged Node.js and Google Cloud to develop cloud functions that streamlined data management and automated manual tasks, boosting daily efficiency and reducing task redundancy by 20% - Led weekly agile team meetings to ensure clear communication and adherence to project timelines

- Identified inefficiencies in nonprofit retreat planning and owned product strategy and KPIs for a centralized budgeting and donation platform, prioritizing budget accuracy and donor continuity. - Led a cross-functional team of 4 engineers and 2 designers to deliver core budgeting and donation workflows, enabling persistent donor tracking and increasing donations by 20%.