Cron data-roll-up to ETL

consumer-facing components

Public APIs

Modified tesseract libraries for transform & OCR

Mild re-training data usage.

Pilot integration for retraining.

CV glue. Modified libraries for pattern-detection. Some pattern training tasks. HoG matching. Transform

Cloudwatch monitoring & events for chron-> lambda.

main data collector -> feeds to ETLs, fires select events.

Low stress ETLs, light master data.

Here & there. All on lambda. Anything guaranteed cold spin-up; otherwise bit slow (compared to node/jvm) when warm.

basic image processing, transform wrapper.

heavy-lifting CV. Re-written C++ libraries for lambda - MSER, text transform, etc.

Public-facing image store.

Too much to list. Long-term, short-term, etc. Some event firing.

cache, ETL activities - roll-up. Anything speed-worthy.

Sparingly. Gateway to non-scaled components.

All critical functions, cron, data-accept -> dynamodb. Light routing.

Internal/admin tools; in-house ground-truth training client.