Sidrun Source | Agentic Procurement for SMBs

How It Works

From need to delivery — fully autonomous

Describe What You Need

Tell Sidrun Source what you want to buy — item, quantity, deadline, and budget. Plain language, no forms.

Agents Handle Everything

Specialist agents find vendors, send quote requests, compare responses, draft a purchase order, and route for approval — all autonomously.

Approve and Track

Review the ranked quote comparison, approve the PO, and track delivery — all in one place. Every decision is logged and timestamped automatically. You decide, the agent executes.

What Sidrun Source Does

Built for lean procurement teams

🔍

Vendor Discovery

Finds relevant, qualified suppliers automatically — no manual searching, no stale vendor lists.

✅

PO & Approval Routing

Drafts purchase orders and routes for manager or VP approval automatically based on order value and policy thresholds.

📊

Quote Comparison

Normalises and ranks vendor responses side by side — price, lead time, terms, all in one view.

📋

Audit Trail

Every sourcing decision logged, timestamped, and explainable. Compliance-ready from day one.

Powered by Sidrun's Intelligence Layer

Sidrun Source is built on top of two standalone tools you can also use independently.

🍋Sidrun Extract

Pull structured data from vendor quotes, contracts, and documents (PDF, DOCX) in seconds.

Prices & Lead Times — Extract key commercial terms from any quote format.
Clause & Condition Extraction — Surface critical terms and obligations instantly.
Structured Output — Clean, tabular data ready to drop into your workflows.
Privacy-First — Documents processed in-browser, never stored on our servers.

Python / Browser Storage Try Extract →

🔒 Privacy-first · No data stored

🍋Sidrun Insight

Transform your supplier contracts and agreements into an interactive, queryable knowledge base.

◎ Coming soon: Knowledge Graph support — structured fact grounding to reduce hallucinations on complex supplier documents.

Vector DB / RAG · Knowledge Graphs Try Insight →

🔒 Privacy-first · Grounded citations

Security & Privacy

Built for sensitive data from day one

🔒

Documents Stay in Your Environment

Your files are never stored on our servers. Processing happens in your browser or your infrastructure — your data stays yours.

🚫

No Model Training on Customer Data

We do not use your documents, queries, or outputs to train or fine-tune any AI model. Ever.

🌐

Browser-Side Processing

Where applicable, extraction and analysis run directly in the browser — no upload required, no network exposure.

📎

Grounded Responses with Citations

AI answers are grounded in your documents and cited back to the source — no hallucinations, no guesswork.

🏢

Enterprise Deployment Options

Deploy on-premise or in your private cloud. Full control over your data pipeline, with no external dependencies.

About Us

Kaidur Roht

Founder · Talk to me →

With over 10 years of experience building document management solutions at Goldman Sachs, Federal Reserve Bank of New York, the New York City Police Department, and the United Nations, Kaidur brings deep enterprise expertise to Sidrun AI — turning hard-won insights from complex, large-scale environments into privacy-focused tools that work for teams of any size.

Procurement that
runs itself.