I need a lightweight proof-of-concept that automatically helps our ServiceNow team resolve new Application-error incidents. Every two minutes the assistant should: • Poll our ServiceNow queue and detect any fresh incident assigned to our group. • Retrieve the relevant time-range logs from Splunk. • Look up similar past incidents stored in our internal database. • Pull any related Slack thread so context isn’t lost. Using that collected data, the tool should run a compact AI model (open-source or API-based—whichever keeps the code simple) to draft concise resolution steps and then post those steps back to the incident record. Key notes – Scope is limited to a working script or small service that proves each integration end-to-end; polished UX isn’t required. – Incidents of interest are strictly Application errors; no need to handle network or security tickets. – Historical records reside in a database, so tap that directly rather than expecting CSV or JSON dumps. Please include links or brief summaries of past work where you connected ServiceNow, Splunk, Slack, or similar platforms. A basic README with setup and scheduling instructions will complete the delivery.