CS 220 · Year 2 · Semester 2 · 3 credits · Computer Science

Software Engineering & DevOps

From Solitary Actuarial Script to Production Risk Engine: Engineering Dependable, Tested, and Deployable Financial Systems.

Full practice exam · 30 Q · 120 min

Season 1 · 8 episodes

Shipping the Engine: DevOps for Actuarial Systems

When a legacy commercial auto rating spreadsheet causes an unhedged underwriting loss, junior actuarial engineer Maya Lin must rebuild the pricing core into an audited, containerized, high-throughput microservice before open enrollment launches.

Protagonist · Maya Lin, a junior actuarial software engineer caught between traditional actuaries who trust spreadsheets and software architects demanding continuous deployment.

Setting · Apex Specialty Underwriters in Chicago, where legacy actuarial reserving tables meet modern cloud infrastructure during a high-stakes multi-state expansion.

Stakes · If the new pricing microservice fails, drifts, or leaks quotes, state insurance commissioners will halt policy issuance across five states, costing Apex twelve million dollars in gross written premium and triggering severe regulatory sanctions.

Pick an episode

Episodes

Season finale: Maya leads the cross-functional team through a flawless canary deployment of the rating engine under full commercial traffic, proving zero mathematical drift, surviving automated chaos injection, and passing the state insurance department audit.

From the lecturer: Welcome to CS 220. Actuarial models do not live in isolated spreadsheets or scratchpads; they power live quotation engines, reserving pipelines, and enterprise risk dashboards that must run with zero downtime under strict regulatory oversight such as ASOP 56 (Modeling). In this course, you will transition from writing individual scripts to engineering collaborative, production-grade microservices. To get the most out of these lecture films, keep your terminal and IDE open beside your video player. Before typing a single line of implementation, draw the system boundary and data flow on paper using the C4 model. When we cover containerisation and CI/CD pipelines, pause the film and reproduce every failure state locally to build diagnostic intuition. While this course does not directly map to a single preliminary SOA/CAS multiple-choice exam, it directly qualifies you for modern actuarial developer roles, predictive modelling engineering tracks, and enterprise risk architecture. Work collaboratively in your assigned git repositories, review your peers' code with intellectual empathy, and treat every test failure as an opportunity to harden your domain logic.

Cold open

At 8:14 AM, Maya arrives at her desk to find three panicked underwriting directors. A broker just bound a fleet policy for two hundred heavy trucks at an eighty percent discount because an actuary updated a territorial modifier in an unversioned Excel workbook on the shared drive.

preparing lecture…

Transcript

Formulate production-ready user stories using the INVEST criteria and Gherkin syntax for actuarial models
Model sprint capacity, flow efficiency, and cycle time using Little's Law
Integrate actuarial model governance (ASOP 56) into Agile Definition of Done (DoD) frameworks