Extraction accuracy
Proportion of fields correctly transferred from proposals to a unified structure.
AI
AI assistant for procurement: proposal analysis and supplier comparison
AI assistant that extracts terms from commercial proposals, compares suppliers by procurement rules, and prepares verified summaries.
Our clients
Clients and partners
How commercial proposals are processed
-
01
Document receipt
The assistant retrieves commercial proposals from an agreed-upon channel: email, form, file storage, or procurement system.
-
02
Terms extraction
The model populates a unified structure: item master, price, delivery term, payment, warranty, and other procurement fields.
-
03
Requirements verification
Each proposal is matched against specifications, thresholds, and mandatory conditions.
-
04
Supplier comparison
System calculates scores using approved methodology and separately shows missing data and deviations
-
05
Solution package
The commission receives a summary, links to source documents, and a list of questions for suppliers.
-
06
Results delivery
After verification, data and documents can be returned to 1C, SAP, Directum, SBIS, or another target system.
Who is responsible for data and decisions
| System / layer | Scope of responsibility |
|---|---|
| Channels and storage | Receiving original proposals and preserving the unchanged original. |
| Extraction module | Field recognition and binding each value to its source |
| Procurement rules | Comparison criteria, weights, thresholds, and mandatory conditions. |
| Procurement commission | Exception verification, clarification requests, and final supplier selection. |
| ERP or DMS | Tracking approved decision, documents, and subsequent operations |
Assess where AI can deliver impact in your process
What needs to be prepared
- An anonymized sample of commercial proposals in various formats.
- Specification or request with mandatory and desired requirements
- Your current supplier evaluation methodology and list of acceptable exceptions.
- An example summary or report that the procurement commission uses.
- A responsible expert who will markup a control sample.
Pilot metrics
First measure the current process, then compare the assistant against the same control sample
Deviation recall
Proportion of known specification mismatches found by the assistant.
Manual corrections
How many values and conclusions the expert changed before the committee's decision
Preparation time
Time from receiving a complete set of commercial proposals to a verified comparative summary.
Pilot
Test the assistant on your proposals
We take an anonymized sample of proposals, align on fields and criteria, compare results against expert markup, and show the rate of manual corrections.
- Real document formats
- Agreed criteria
- Expert control sample
- Error report
Discuss solution: AI assistant for procurement: proposal analysis and…
Related Cases
OSNO-VA: AI accountant
- Built an AI platform
AI ingredient recognition by barcode
- Processing sped up from 30 minutes to 2 per batch of 10 images
- Recognition accuracy for composition is 80-95%
OSNO-VA: Financial Agent on MCP
Learn moreRelated solutions
Articles on the topic
- AI Agents and Federal Law GDPR: How Not to Expose Personal Data to an LLM
- How to Choose an AI Implementation Contractor: Questions and Red Flags
- AI Accountant for a Retail Chain: How to Automate Revenue Reconciliation
- How much an AI agent costs: what determines price and TCO
- The illusion of parity: open AI models vs the frontier
