Improving Workflows for LinkedIn Outreach and Tech Stack Identification

Hi everyone, I’m looking for help with two workflows and would really appreciate any insights:

I'm building a workflow to determine if a prospect is active on LinkedIn, so we can decide whether to include them in our linkedin outreach. While I’m able to enrich profiles and check if they're active, the issue I’m facing is that Clay doesn’t seem to visit the profile to check for recent engagement, such as likes, comments, or posts. Clay retrieves most data points from profiles, but the results for activity aren’t accurate. Currently, I’m using GPT-4 mini and our own OpenAI API for this. Could someone help me figure out a better approach? 2. Identifying tech stacks for construction companies: I'm also trying to identify specific tech stacks used by construction companies. I’ve attempted this using Clay’s tech stack waterfall enrichments but only achieved a success rate of 9.1%. I’ve also tried scraping job descriptions via Clay to extract tech stacks but haven't had much luck. Any ideas on how to work around this or improve accuracy? Any help would be greatly appreciated! Thanks in advance!