Building data-intensive backends, distributed crawling platforms, and AI-assisted data systems across B2B intelligence workflows. Focused on the infrastructure layer: high-throughput crawling, geospatial entity resolution, cloud-native pipelines, browser automation, and LLM tooling that turns raw web data into operational products.
Browser-grade HTTP SDK and CLI for crawling infrastructure. Wraps curl_cffi with TLS impersonation, typed responses, proxy rotation, bounded async concurrency, retry policy, and terminal debugging workflows.
Controlled distributed HTTP task proxy where workers connect outbound to a Mastermind service over mTLS. Built around gRPC bidirectional streaming, safe outbound dialers, SSRF hardening, task epochs, and backpressure.
Product-intelligence matching engine that extracts structured identity from marketplace pages, orchestrates browser-backed competitor search, and scores candidates with category-aware entity resolution.
Next.js decision platform where role-based LLM agents debate a thesis under an active moderator. Prisma models sessions, objections, risk clusters, telemetry, and final decisions as an auditable record.