Quote Requests (RFQs) and AI Comparison

Sourcing the same bill of materials from several suppliers usually means juggling email threads, attachments, and inconsistent quote formats. Streckenheld AI turns that into a single workspace: one quote request, many recipients, every offer in one table, and an AI-generated category-by-category comparison ready to act on.

Key features

• One BOM, many recipients — upload a bill of materials once and send it to as many suppliers as you need
• Storefronts and contacts — pick from your featured retailers or any contact in your address book; emails are pre-filled automatically
• Free-text recipients — add a supplier name with no record yet; their offer still lands in the same table
• Track offer status per recipient: pending, sent, offer received, accepted, declined
• Add offers received outside the system — drop in a PDF you got by email or WhatsApp and attach it as a new offer with its own supplier name
• Inline preview / download of every offer file (and the original BOM) without leaving the page
• Expand a row to inspect the line items the AI extracted from the offer
• AI comparison with thinking — the model streams its reasoning live and produces a markdown report grouping items by inferred category, picking a recommended supplier per category, and a final overall recommendation
• Custom comparison instructions — a per-RFQ textarea (e.g. "prioritize delivery speed", "ignore offers without insulation") that the AI follows when comparing
• CSV download on every comparison table — one click to export any breakdown to spreadsheet
• Soft-delete offers with a single trash button so the comparison only sees the ones you care about
• Accepted offers become contracts — marking an offer accepted creates a contract for the construction site referencing the same offer file and kicks off the standard contract analysis (catalog extraction, price/payment-term checks)

How it works

1. Create the quote request. Upload your bill of materials (BOM) as a PDF and pick the construction site. The AI runs the OCR/extraction in the background while you set up recipients.
2. Add recipients. Tap a featured storefront, pick a contact from the supplier combobox (free-text supported), or paste an email manually. Name and email auto-fill from the contact / saved retailer email.
3. Send the request. Each recipient gets a personalized email with the BOM attached and a link back to the request.
4. Collect offers. Suppliers reply by email; you upload their PDF directly to the row. For offers that came in through other channels (e.g. WhatsApp, a phone call follow-up), use Add offer to drop the file in with an optional supplier name and email.
5. Watch the AI extract line items. Each new offer is OCR'd and indexed; expand the row to see the parsed quantities, units, and prices.
6. Write your comparison instructions. Inside the Quote request card, give the AI any guidance specific to this RFQ — preferred suppliers, weighting price vs. delivery, exclusions.
7. Generate (or regenerate) the AI comparison. The model streams its thinking and the markdown report appears live: per-category tables, recommended supplier per category, and an overall comparison with a one-line recommendation.
8. Accept the winners. Accepting an offer files it as a contract for that construction site and starts the catalog/price-check pipeline so future invoices from that supplier are automatically validated against the agreed prices.

Who it's for

Quote Requests are built for procurement managers and project leads who need to compare supplier offers fairly and quickly — whether the offers came through the formal request flow or off-channel. The AI handles the boring "line up the columns" work so you can focus on the decision.