Delivery lesson · ACT-IAC AI Hackathon
The deadline was part of the outcome.
The 13-day limit forced a four-person team to decide what had to work, how it would be tested, and what a judge could inspect. The Integrity case study has the full technical record.
What Happened in 13 Days
A four-person Intelligrit team had 13 days to build and present Integrity for the ACT-IAC 2026 AI Hackathon. The team delivered a judged MVP that processed public CMS data, produced inspectable provider-risk signals, and took first place in the CMS track.
The Integrity case study has the metrics, architecture, validation method, provenance, and limitations. This note focuses on what the fixed delivery window changed about the work.
What the Deadline Changed
The team had to decide what a judge could actually use and inspect. That pushed effort toward a repeatable data pipeline, a working investigation interface, measurable validation, clear deployment instructions, and an operating footprint that could be explained.
The team narrowed
- A small, explicit definition of done
- Evidence selected before the final demonstration
- Architecture proportional to the evaluated need
- Fast decisions about what not to build
The team still had to prove
- Objective validation and known limitations
- Human review of analytical signals
- Deployment and operating clarity
- Accountability for what was presented
If Speed Matters, Specify It
“Fast” is too vague to buy. Name the capability, due date, evidence to be reviewed, safeguards, and official who can decide. That turns schedule into measurable performance.
A shorter window can raise price, staffing pressure, dependency risk, and the need for timely Government decisions. Those tradeoffs belong in the requirement, and every required gate still has to fit inside the agreed window.
What This Result Supports
This was a bounded hackathon using public data, a known environment, judges, and a fixed deadline. Production authorization, current-data performance, operational fit, accessibility conformance, and agency acceptance would each require separate evidence.
The result supports a narrower lesson: a clear deadline and evaluation method helped a compact team spend its time on something a decision-maker could inspect.