Tips and best practices

Before analysis

Upload all RFP documents before adding bids. The AI uses these as the baseline for every check.
Set up evaluation criteria before your first analysis. Adding criteria after and re-running costs double.
Wait for parsing to complete on all documents. Running analysis on partially parsed documents gives incomplete results.
Use clear names for procurements and bids. "IT Infrastructure Upgrade 2026" and "SIA TechnoGroup" are easier to work with than "Procurement 1" and "Bid A".
Check document quality. Clear scans, unlocked files, native digital formats when available.

Don't close your browser if you want to watch progress. But if you do, the analysis continues server-side.
Check for clarification requests periodically if you're not watching live. The analysis pauses until you respond.
Let it finish. There's no cancel button. The analysis needs to complete to produce reliable results.

Start with the strategic briefing. It gives you the big picture in 30 seconds.
Focus on Critical and Major findings first. These are the ones that matter most. Minor findings and notes can wait.
Always check the evidence. The AI's conclusion is only as good as its interpretation. The quotes let you judge for yourself.
Don't panic over finding counts. Ten Minor findings are less concerning than one Critical. Severity matters more than quantity.
Look at Strengths too. They're useful for evaluating the overall quality of a vendor's response, not just the problems.

Analyze all bids before comparing. Running analyses at different times with different criteria makes comparison unreliable.
Use evaluation criteria. Comparison without criteria is based on concern counts alone, which is less informative.
Export for your team. The Excel export captures everything in a shareable format.

Prepare documents well. Clean, complete uploads mean better results on the first run. Fewer re-runs = fewer credits spent.
Remove irrelevant pages. That 100-page appendix of general terms that applies to every procurement? The AI will read it and charge you for it.
Configure criteria once, upfront. Don't add criteria one at a time and re-run each time.
Don't re-run just because you don't like a finding. Check the evidence first. The same documents will produce similar results.

The AI is a tool, not a decision-maker. It catches things humans miss in long documents, but it doesn't have context you do. Use its output as input to your judgment.
Upload everything the vendor sent. Supporting documents, CVs, certificates - they all contribute to a more complete analysis.
Export early, export often. Keep copies of your analysis results outside the platform, especially before deleting procurements.