Synthetic data
The Meridian Re datasets the examples read: what each file holds and which chapter uses it. All of it is synthetic, generated from a fixed seed, and none of it is real.
data/ is generated deterministically by generate_data.py from a fixed seed, ensuring byte-for-byte reproducibility on any fresh clone. No proprietary company data or real mortality tables appear anywhere in this work.Motor India paid and reported loss triangle, accident years 2018 to 2023, development 12 to 72 months, cumulative USD. Four defects are seeded for the data-quality agent to find.
Note: Defects: two blank cells, one negative case reserve, one row where reported is below paid, and one understated claim count.
Used by
The same triangle without the seeded defects, for the reserving workflows.
Used by
A Gompertz-shaped synthetic mortality table with the structure of IALM 2012-14 ULP: ages 18 to 99, by gender and smoker status. Not the real table; do not use it for real work.
Used by
Sixty comparable commercial property accounts with total insured value, premium and five-year incurred losses, for underwriting benchmarks.
Used by
A synthetic broker submission for a Chicago commercial property risk, read by the COPE extraction tool.
Used by
Five hundred term life policies: age at entry, sum assured in INR, term, premium frequency, smoker status, lapse indicator.
Note: Generated for the Chapter 12 narrative; no script reads it yet.
Used by
Three quarterly experience-study notes for term life India, the archive the Chapter 12 vector-knowledge agent indexes.
Used by
Three hundred UK annuity members: date of birth, annual pension in GBP, commencement date, level or escalating, dependant indicator.
Used by
Two capital-model snapshots, FY2025 Q1 and Q2, with six SCR modules and the parameter versions behind each.
Used by
Monitoring thresholds and seven-day metrics for two deployed agents: tool error rate, tool calls per run, p95 latency, escalation rate, schema failures, cost per run.
Used by