
Joshua A
Claude Code, Claude Cowork Git Automation and Workflow Generation
Compétences

Voir mes services


Expérience professionnelle
Claude Cowork automation systems — finance ops via desktop AI
Finance Expert • Freelance
Jan 2026 - Mar 2026 • 2 mos
The finance team handled large volumes of documents manually, reconciliations, reports, and file organization were slow and error-prone. I deployed Claude Cowork to automate multi-step desktop workflows, allowing the system to process files, generate reports, and manage data directly within their environment. I’ll be honest, fine-tuning how it handled edge-case documents took a few iterations, but once stable, it performed consistently. In most cases, it saved 18+ hours weekly, improved accuracy, and streamlined financial operations without requiring technical expertise from the team.
Claude Cowork desktop automation — operations via MCP connectors
Automation Solutions • Temps partiel
Oct 2025 - Nov 2025 • 1 mo
The operations team relied heavily on manual file handling, reporting, and repetitive desktop tasks that drained time and introduced errors. I deployed Claude Cowork as a desktop automation agent, enabling direct interaction with local applications to manage files, generate reports, and execute multi-step workflows without manual intervention. I’ll admit, the first iteration needed refinement around edge-case file handling once fixed, it ran clean. In most cases, it saved 15+ hours weekly, reduced operational friction, and allowed non-technical staff to automate complex workflows without touching code.
Claude Code AI agent systems — software dev via GitHub
Software Keys Hub • Temps plein
Sep 2025 - Oct 2025 • 1 mo
Developers were spending too much time on repetitive coding tasks and manual debugging, slowing overall product velocity. I integrated Claude Code into their workflow, enabling automated code generation, testing, and debugging directly within the terminal while maintaining structured GitHub pipelines. There was a late-night test where the agent resolved a recurring bug faster than expected that changed internal trust immediately. In most cases, it accelerated development cycles, reduced manual workload, and improved engineering efficiency across the team.