Arun Kumar

Owning the full data collection and delivery lifecycle — engineering scrapers that handle bot defenses, building ETL pipelines, and delivering clean, warehouse-ready datasets for business consumption.

• ▹Built production-grade scrapers in Python using Requests, BeautifulSoup, Playwright, and Apify — covering static HTML pages through to fully dynamic, JavaScript-rendered platforms including Pinterest.
• ▹Designed multi-stage ETL pipelines that normalize raw scraped data, resolve duplicates, and enforce schema contracts before loading into the data warehouse.
• ▹Automated end-to-end data workflows via Python scheduling — eliminating manual extraction runs and reducing data delivery to near-zero human intervention.
• ▹Implemented transformation logic to handle inconsistent formats, null fields, and nested JSON structures across heterogeneous source sites.
• ▹Integrated processed datasets with Metabase to power reporting dashboards and business intelligence queries.
• ▹Built resilient scraping infrastructure with retry logic, request throttling, and session management to sustain throughput against anti-bot systems.

Building core backend infrastructure for a US-based fintech startup — REST API design, database architecture, authentication systems, and banking workflow engineering.

• ▹Architected and shipped RESTful APIs with FastAPI powering core banking workflows: account management, transaction processing, and user onboarding.
• ▹Designed and enforced JWT-based authentication and role-based access control (RBAC) across all protected endpoints.
• ▹Modeled relational database schemas in PostgreSQL optimized for financial data integrity, referential consistency, and concurrent-safe operations.
• ▹Engineered backend workflows for fintech-specific features — fund transfers, balance ledgers, and statement generation.
• ▹Implemented Redis caching for high-frequency API responses and session data, reducing database load on hot paths.
• ▹Containerized backend services with Docker for consistent local development and production deployment environments.
• ▹Built and maintained a double-entry ledger system for tracking financial transactions with full audit trail support.

• ▹Integrated ML models into production applications, building inference wrappers and API layers to expose model outputs as usable product features.
• ▹Profiled and optimized model inference pipelines to reduce prediction latency for real-time use cases.
• ▹Delivered AI-powered features across cross-functional teams, translating research outputs into stable, deployable backend services.

• ▹Designed and implemented multi-step agentic AI workflows using LLM orchestration for automated UX research tooling.
• ▹Built automation pipelines that replaced manual UX research tasks through AI-driven data extraction and synthesis.

• ▹Built pixel-perfect, responsive landing pages and portfolio sites in React (TypeScript) with Tailwind CSS.
• ▹Implemented component-driven UI architecture ensuring cross-browser compatibility and consistent design fidelity.

• ▹Delivering custom data engineering, scraping pipelines, and backend solutions for international clients across e-commerce, research, and analytics domains.
• ▹Scoped, architected, and shipped complete client projects end-to-end — from requirements to deployment.