Agentic Problem Solving Team

What if you could build a Swahili dictionary, file your own bug reports, and write your own code fixes — all while you're away from your desk? In this episode of Practical AI for Kingdom Work, Matthew walks through how he built a team of specialized AI agents inside Claude Code that work together to tackle complex language data tasks, verify their own results, and commit clean code to GitHub. WHAT WE COVER IN THIS EPISODE: 🔹 Agent roles and how to define them — how Matthew set up a team lead, programmer, quality control, domain expert, simplify bot, and archivist using simple markdown text files 🔹 Building a Swahili lexicon autonomously — how the agent team created over 2,000 linked dictionary entries in FLEx, encountered bugs, wrote workarounds, and filed GitHub issues for later resolution 🔹 Managing agents from your phone — two tools for approving, denying, or redirecting agent tasks remotely so your computer keeps working while you're away RESOURCES MENTIONED 🔹 Claude Code: https://www.anthropic.com/product/cla... 🔹 Happy Engineering (remote mobile access for Claude Code): https://happy.engineering 🔹 FLEx (FieldWorks Language Explorer): https://software.sil.org/fieldworks 🔹 GitHub: https://github.com Learn more about our language tech initiatives: https://software.sil.org Is your organization a nonprofit? You may qualify for up to 75% off Claude: https://claude.com/solutions/nonprofits ━━━━━━━━━━━━━━━━━━━━━ ABOUT PRACTICAL AI FOR KINGDOM WORK Welcome to the intersection of faith, language, and technology. We explore how artificial intelligence and language technology can practically support Bible translation, scripture engagement, and the flourishing of language communities around the world — equipping you to use these tools in actionable, ethical ways. 👉 Explore the full episode library: https://podcast.langtech.cloud/ #FaithAndTech #KingdomWork #LanguageTech #SIL #ArtificialIntelligence #MinistryTools