1. Home
2. Help
3. Getting started
4. Your first analysis

Your first analysis

Sign up, create a project, upload a document, run an engine, read the result. Ten minutes, one concrete output — and you'll understand how the rest of Trueleveler works.

Before you start

You need a Trueleveler account (the free 14-day trial is enough) and one document to analyze. If you don't have one handy, skip to sample data — every engine ships with a pre-loaded example.

Step by step

1. Sign up or sign in

   Head to trueleveler.com/login. If you already have an account, sign in with email + password or via Google/Microsoft OAuth. If you're new, click Create account — you'll be on a 14-day Pro trial with 15 AI runs and access to all engines.

2. Open the app

   After sign-in you land on /app. The left sidebar is your main navigation: Projects, Documents, Vendors, Engines, Trackers, Analytics. The top bar shows your active project and your remaining runs this month.

3. Create a project

   Click Projects in the sidebar → + New project. Give it a name (anything works — "Bridgeport HQ" or "Downtown Office"). This creates a container for documents, vendor records, trackers, and engine sessions scoped to one job.

   Everything you do in Trueleveler is scoped to a project unless you explicitly view across projects (e.g., in Analytics).

4. Upload a document

   Click Documents in the sidebar. Drag your file into the drop zone, or click to browse. Supported formats: PDF, DOCX, XLSX/XLS, CSV, PNG/JPG.

   You don't have to upload from Documents — most engines accept uploads directly when you start a run — but uploading here first means the document is reusable across engines.

5. Pick an engine

   Click Engines. You'll see 19 cards grouped by project phase: Pre-Award, Award, Post-Award, Field, Tracking. Each card has a short tagline and an approximate runtime.

   If you uploaded a contract, click Contract Review. A bid? Bid Leveling. A pay application? Pay App Review. Not sure which one? Most cards have a Try with sample data chip — click it to see the output without using a run.

6. Attach the document and run

   The engine's input panel tells you what documents it needs. For single-doc engines (contract, scope, pay app) just pick the doc you uploaded. For multi-doc engines (Bid Leveling, PO vs Invoice) you'll attach one doc per vendor or one pair.

   Click Run analysis. The progress bar shows the model thinking — most engines finish in 20–60 seconds.

7. Read the results

   Every engine opens its results in the same layout:

   • Header: engine name, confidence score, export buttons
   • Summary: 2–3 sentence AI narrative
   • Detail: structured output — line items, risk flags, tables, whatever that engine produces
   • Chain suggestions: what to run next (e.g., Contract Review → Insurance Tracker → PM Handoff)
   • Chat: at the bottom, ask questions about this specific result

8. Export or chain

   At the top right of the result you can export to PDF, XLSX, or CSV. Or click one of the chain buttons at the bottom to run a follow-on engine — most carry the relevant data forward automatically so you don't re-upload.

What just happened

The engine sent your document to a large language model (Claude or Gemini, depending on the engine) with a carefully structured prompt that told the model what to look for, what format to return, and how to flag risks. The output was parsed into structured JSON and rendered in the results view.

The session is saved to your project automatically. You can find it later under Results or History in the sidebar. Your team (if you've invited anyone) can see it too.

What to do next

• Try a second engine — many chain directly from the first. Bid Leveling → Contract Review is the most common one.
• Add vendors to the Vendor Database so engines can tag scoring automatically.
• Invite your team from the Team page — they see the same projects and results.
• Browse the full engine list in the sidebar to see what else is available.