Engineering Status
Quarterly goals · Q4 2026 · October to December
Accuracy & LLM Cost
90.7%
grading accuracy · all configs excl. beta · 104,516 teacher-reviewed rubric evals since Oct 1
LLM cost / activity · October 2026
$0.81
Papers needing manual review
24.0%
Targets
0/3 set
Target: Grading accuracy — not set · LLM cost per activity — not set · Papers needing manual review — not set
View App Speed
117s
processor p95 for 30–50-page uploads · 2,839 of 13,203 runs since Oct 1
Median
55s
Runs within 2 min
95%
User-facing app
audit pending
Target: processor < 2 min per 40 papers (p95) · app < 2 s to load
View Documentation
—
A new dev ships a small change in week one
Scope
Setup · Rules · Architecture · AI context
Owners
one per repo
Target: a new dev from a different department can onboard on their own
View Scalability
—
No stress tests recorded yet
Concurrent users
— / 50,000
PDFs in one go
— / 100
Concurrent teacher scans
— / 125
Target: 50,000 concurrent users · 100 PDFs in one go · 125 concurrent scans, without degradation
View