Government Procurement Intelligence

Problem It Solves

In government procurement, bidders often need to track scattered public notices, eligibility rules, scoring criteria, deposit requirements, small-business price adjustments, contract clauses, competitor history, and compliance risks. The information is fragmented, deadlines are tight, and PDF or Word bid documents are easy to misread. This skill turns that workflow into a concrete analysis pipeline: finding projects, judging whether to bid, estimating pricing risk, and surfacing bid-document issues.

How It Works

• Opportunity discovery: searches public procurement sources by industry, region, budget, and procurement method, then returns a ranked daily digest.
• Bid decision support: checks qualification requirements, competitor profiles, and scoring rules to provide pricing guidance and risk notes.
• Compliance auditing: supports collusion and bid-rigging checks, timing compliance, scoring anomalies, contact-overlap checks, and relationship mapping.
• Document support: assists with tender-file parsing, technical proposal drafting, qualification filling, case reference, format checks, and reusable material libraries.
• Calculations and drafting: handles bid deposits, small-business price deductions, objection or rejected-bid appeals, contract-clause risk, and cross-province policy differences.

The usual flow is to configure the company profile first, such as industry, qualifications, region, and budget preference, then provide a project link or keyword query. The skill combines public data with rule-based or model-assisted analysis and labels low-confidence results when data is incomplete.

Boundaries

It is useful for common procurement categories such as IT services, office equipment, construction, medical supplies, and property management. PDF parsing, Word/PDF export, OCR, and graph analysis may be unavailable or degraded if tools like pdfplumber, python-docx, jinja2, or OCR are missing. Scoring prediction and competitor profiling improve with accumulated historical data; without data or an available LLM, the output falls back to rule-based estimates. Anti-crawling controls, login walls, and network timeouts can limit collection, so official notices and tender documents remain the final authority.