● Building a 0→1 AI-assisted audio analysis product for music producers, owning product development end-to-end across frontend, backend, infrastructure, and UX. ● Built the application with Next.js, React, and TypeScript, including authentication, audio upload, analysis workflows, and interactive visualization of track metrics. ● Built backend services and APIs for audio processing and analysis, transforming uploaded tracks into structured metrics consumed by the frontend. ● Designed the product's data and storage architecture using Firebase, Firestore, and Cloud Storage, including user-scoped access controls, file validation, and secure upload workflows. ● Building AI-assisted analysis workflows that use structured audio metrics as context to generate actionable feedback for music producers.

● Reduced API onboarding time by ~40% by building self-service workflows and frontend tooling around ~800 API endpoints for a developer platform serving 200+ enterprise clients. ● Reduced frontend build time from ~15 minutes to 1–2 minutes, substantially shortening the development feedback loop and improving iteration speed. ● Owned frontend architecture and delivery end-to-end as the sole frontend engineer in a cross-functional team, from ambiguous product requirements and API design through implementation, testing, launch, and iteration. ● Reduced recurring integration friction for enterprise API consumers by turning direct user feedback and common support problems into self-service product workflows. ● Built a static API documentation site with Next.js, using static generation to deliver developer-facing API documentation. ● Improved API integration workflows by partnering directly with backend engineers on API contracts, validation requirements, and platform capabilities. ● Expanded team capacity by mentoring an intern into an independent contributor, providing technical guidance, code reviews, and ownership of production work.

● Removed developers from routine smart-contract validation workflows by building a self-service testing product for QA engineers from scratch. ● Enabled non-engineering teams to update production content independently by integrating a headless CMS into user-facing Next.js applications. ● Reduced duplicated frontend implementation across products by contributing reusable React components and complex form primitives to Lido's shared UI kit. ● Maintained and extended production Next.js applications, shipping user-facing functionality within an established frontend architecture. ● Improved consistency of complex input flows through reusable form architecture, validation patterns, and shared UI components.

● Shipped production frontend functionality to tens of millions of users on Mail.ru, working on a high-traffic consumer product. ● Supported critical authentication and account flows at scale, including VK ID integration where reliability and backwards compatibility were essential. ● Stabilized a test suite containing 1,000+ tests, reducing CI-related friction and improving confidence in production releases. ● Improved production reliability for a high-traffic consumer product through on-call support, incident investigation, debugging, and production fixes.

● Eliminated developer dependency from a manual content-publishing workflow by designing and building a self-service CMS for internal content teams. ● Replaced script- and command-line-based publishing with a browser-based product designed around the day-to-day workflows of content managers. ● Enabled content teams to create and manage reusable email content independently by replacing developer-managed publishing with reusable templates. ● Owned the product from problem discovery through production delivery, working directly with users, defining requirements, coordinating backend API changes, and implementing the frontend. ● Improved the product after launch through direct user feedback, turning operational issues into iterative frontend and workflow improvements.

● Improved frontend performance using Core Web Vitals and direct user feedback to identify bottlenecks, then applying lazy loading to reduce unnecessary upfront loading. ● Improved response times for blocking user flows by working with backend engineers to identify slow requests and optimize backend operations affecting frontend performance. ● Established production performance monitoring with Dynatrace, giving the team visibility into real-user performance, bottlenecks, and regressions. ● Introduced GraphQL into a digital-signature module from scratch, integrating Apollo and establishing GraphQL-based communication between the frontend and backend. ● Improved the platform through direct user feedback, gathering feedback during regular demos and converting recurring performance and usability problems into engineering improvements. ● Improved frontend reliability and developer experience through reusable UI components, shared validation primitives, automated testing, and common platform tooling.

● Increased conversion by ~15% by improving customer-facing tariff customization flows and iterating on the user experience using product data. ● Shipped frontend functionality for a telecom product serving 47M+ users, including core tariff selection and customization journeys. ● Enabled data-driven product decisions by introducing product analytics and A/B testing into customer-facing frontend flows. ● Improved call-center operational workflows by simplifying frequently used CRM interactions for internal operators.