AI Risk Scoring for Indian Civil Construction Tenders
Score the risk on a 500-page tender in 20 minutes, not 4 days
Civil Brain reads the entire tender, scores five weighted dimensions, and returns a 0-100 health score with one of four bands: Low Risk, Moderate, Needs Attention, or High Risk. Every clause that triggered a flag is linked by page number, not summarised away.
The bid you wish you had skipped costs more than the one you lost
Tender risk-reading is a senior-team activity. A junior engineer skims the BOQ; the senior partner reads the special conditions. The general conditions, which often run 200 pages, get glossed. The LD clause, the defect liability period, the mobilisation advance terms, the force majeure scope. These are the clauses that bite at month two.
Risk reading at the bid stage is cheap. Risk discovery at month two is expensive.
General conditions are skipped, not read
Most bid teams read the BOQ and the special conditions, then bid. The general conditions, where 80% of risk-shifting clauses live, get a five-minute skim. The clauses that do not deviate from FIDIC are fine. The clauses that do are not.
Standard forms are not standard
FIDIC Red, FIDIC Yellow, MoRTH ITB, state PWD General Specifications. Each defaults differently on retention, defect liability, mobilisation advance, and force majeure scope. Reading them as if they are identical is how surprises happen.
LD clauses appear in three places
Liquidated damages typically appear in the special conditions, the BOQ preamble, and sometimes the technical specs. Each can specify a different LD rate. The number that wins is usually the one that hurts.
Five factors, one health score, every clause traced
Civil Brain reads the tender end to end. The health score is a weighted composite across five dimensions. Click any factor to see the clause that moved it.
Risks (40% weight): critical to low, clause by clause
Every identified risk is classified by severity: critical, high, medium, or low. Each card shows the clause reference, the financial impact estimate, and the recommended mitigation strategy. Critical risks drop the score steeply; low risks move it minimally.
Compliance (20% weight): 27-item checklist, not a summary
Every compliance requirement in the tender is extracted and tagged: verified, to-verify, or failed. Covers eligibility criteria, submission obligations, and mandatory conditions. Failed items drop the compliance sub-score sharply and surface at the top of the risk report.
Clause Conflicts (15% weight): Volume II vs Volume III
Civil Brain cross-references all uploaded volumes and flags contradictions between them. An LD cap specified at 10% in the special conditions and 20% in the general conditions is a conflict. Each conflict surfaces with the two clauses side by side.
Financial Exposure and Timeline (25% combined)
Financial Exposure totals estimated liability across penalties, guarantees, and insurance requirements. Timeline tracks critical dates and flags anything overdue or within 14 days. Both feed into the health score and the bid decision summary.
How risk scoring runs
Most bid teams run the risk score before deciding whether to assign a senior reader to the tender at all. The health score and bid verdict replace the gut call.
Upload the tender, get the risk read in 20 minutes.
Upload all volumes
BOQ, special conditions, general conditions, technical specs, drawings. Drop them all in. Civil Brain handles 500-1,500 page tenders across multiple PDFs.
Civil Brain reads end to end in 20 minutes
Five pipeline stages: structure mapping, extraction, cross-referencing, risk assessment, and report compilation. Cross-references special conditions against general conditions, finds contradictions, identifies deviations from MoRTH or FIDIC defaults.
Health score with clause traces and bid verdict
The 0-100 health score is a weighted composite across five factors. 80+ is Low Risk. 60-79 is Moderate. 40-59 is Needs Attention. Below 40 is High Risk. Every risk card links to the specific subclause and page number that triggered it, with a one-line explanation of the deviation from norm.
Inside the tender intelligence workspace
Risk Scoring ships alongside the rest of the Tender Intelligence workspace. They share the same uploaded document, the same project record, the same Civil Brain reading across all of them.
Run risk scoring on a tender you are considering.
30 minutes with a construction expert. Send a tender PDF you are evaluating; we will run risk scoring on the demo call and walk you through every red flag.