{"id":629,"date":"2025-11-24T20:43:25","date_gmt":"2025-11-24T20:43:25","guid":{"rendered":"https:\/\/camilo.matajira.com\/?p=629"},"modified":"2025-11-25T20:29:40","modified_gmt":"2025-11-25T20:29:40","slug":"summarizing-public-information-of-current_company-using-aws-bedrock","status":"publish","type":"post","link":"https:\/\/camilo.matajira.com\/?p=629","title":{"rendered":"LLM Agent based on public information of ${CURRENT_COMPANY} using AWS’ Bedrock"},"content":{"rendered":"\n

Summary<\/h2>\n\n\n\n

As part of a AI hackathon, I created a simple tool that allowed me to query with human language the public documentation of the enterprise I work in. I leveraged mainly AWS Bedrock, with a Knowledge Base and a single Agent. The results are promising, given that the data was not heavily ‘massaged’ due to time constraints.<\/p>\n\n\n\n

Data wrangling<\/h2>\n\n\n\n

To download all public documentation of Powens, I used wget<\/code>.<\/p>\n\n\n\n

<\/circle><\/circle><\/circle><\/g><\/svg><\/span>