Nikhil Bhardwaj

Store Backend Platform

Joined pre-IPO and grew with the company from a small team into the engineering org behind its e-commerce business — the custom platform that sells and fulfills prescription and OTC products directly, rather than bolting onto an off-the-shelf storefront. Progressed from Senior to Tech Lead to Staff Engineer as both the team and the scope of my ownership expanded.

• Grew the in-house subscriptions platform from a single-table, pay-as-you-go prototype into a configurable system serving ~3.6M billing-active subscriptions. Drove the evolution from tightly-coupled billing to fully decoupled payments and shipments, and expanded from the original 3 treatment categories to the full catalog — including the higher-value Rx categories (Weight Management chief among them) that became the company’s biggest growth engine. Partnered closely with newer engineers throughout, sharing the reasoning behind the design and getting them up to speed as owners of the system.
• Led the identity and access-management overhaul, including a migration of ~23M user accounts to Auth0 delivered ahead of schedule. Owned the design and use-case analysis and coordinated across the Growth frontend, Cybersecurity, and Auth0 vendor teams to move core login/signup off a legacy database-backed system with no regression in login-success or signup-conversion rates — then extended the same platform with adaptive multi-factor authentication, ReCaptcha bot protection (tuned to cut false-positive challenges), and SMS-based passwordless login to harden it against account takeover and abuse.
• Built and scaled the order-tracking system, expanding it from basic one-directional shipment tracking to bi-directional tracking for at-home lab kits (both the outbound kit and the customer’s return sample), and extending the same model to in-person lab visits. Designed it for extensibility so downstream teams could build on a shared foundation instead of one-off point solutions.
• Handled capacity and scaling planning for peak-traffic events, including preparing platform infrastructure for a Super Bowl TV-ad traffic spike.
• Delivered across new business lines as the company expanded — Mental Health (psychiatry and therapy), Skincare, GLP-1, and others — repeatedly turning vague requirements and tight timelines into shipped launches, and authored platform-improvement proposals spanning product end-of-life, Visits, and the Product Catalog API.
• Automated dead-letter-queue management using AI-assisted development. Turned a recurring manual on-call chore into a deployed, largely autonomous pipeline that resolves known failure modes on its own and surfaces only novel cases for human review. Proficient with Claude Code and multi-agent orchestration workflows, and focused on using them to eliminate operational toil rather than as one-off experiments.

Worked in the Delivery Experience org to unify and present accurate, consistent delivery information to customers across the shopping journey — from search and detail page through cart and checkout. Later moved into Softlines, building systems to take the guesswork out of buying clothes and shoes and help customers get the right size and fit. Also mentored interns and new engineers.

Relocated from India to the US to work on the RESTful APIs powering Audible’s native mobile apps.

• Led development of content delivery and streaming for Amazon English, an immersive experience for teaching English to intermediate learners through authentic spoken content.
• Built a content-to-content similarity algorithm to power relevant content recommendations for listeners.

Started my career on the Transaction Risk / Fraud Detection and Abuse Prevention team, building online services within a large service-oriented architecture.

• Built the near-real-time data integration behind the Concessions Abuse Prevention program — a launch with ~$100M in annual impact. Integrated the Variable Computation Library (VCL) precompute store into the Concessions Abuse Service, replacing a pipeline that depended on stale data warehouse feeds with customer-level order-history aggregates available in near real time — and built the workflow that caught abuse of free-replacement orders. Reused Buyer and Digital Fraud team components to widen overall abuse capture.
• Built a monitoring dashboard that cut the team’s operational load — a central place to manage all monitors and alarms, with one-click monitoring setup for launching a new service stack from an existing one.