Auto Ingestion Console
Batch transcript ingestion — NVIDIA LLM evaluation pipeline
llama-3.3-70b-instruct 2026-06-07 22:44 UTC
ON
Scheduler
753
Total Processed
0
Total Errors
2026-06-07 22:39:06
Last Run
Transcript Source Folder
Place your .txt transcript files in this folder. The agent checks for new files on each run.
Manual Batch Run

Trigger a single ingestion run immediately. New transcript files will be evaluated and added to the call register.

Automatic Scheduler
Scheduler is currently ENABLED.
Pipeline Configuration
Batch Size30 files/run
Rate Limit36 calls/min (free tier safe)
Max throughput~51840 calls/day
LLM Modelllama-3.3-70b-instruct
Routing Policyboth
Ingestion Run History Last 27 runs
Run Time (UTC) Found Done Failed Details
2026-06-07 22:39:06 30 30 0 30 call(s) evaluated
2026-06-07 22:32:39 30 30 0 30 call(s) evaluated
2026-06-07 22:28:02 30 30 0 30 call(s) evaluated
2026-06-07 22:22:50 30 30 0 30 call(s) evaluated
2026-06-07 22:19:30 30 30 0 30 call(s) evaluated
2026-06-07 22:15:00 30 30 0 30 call(s) evaluated
2026-06-07 22:10:51 30 30 0 30 call(s) evaluated
2026-06-07 22:06:43 30 30 0 30 call(s) evaluated
2026-06-07 22:03:04 30 30 0 30 call(s) evaluated
2026-06-07 21:58:13 30 30 0 30 call(s) evaluated
2026-06-07 21:53:23 30 30 0 30 call(s) evaluated
2026-06-07 21:49:05 30 30 0 30 call(s) evaluated
2026-06-07 21:43:36 30 30 0 30 call(s) evaluated
2026-06-07 21:39:37 30 30 0 30 call(s) evaluated
2026-06-07 21:34:42 30 30 0 30 call(s) evaluated
2026-06-07 21:30:00 30 30 0 30 call(s) evaluated
2026-06-07 21:26:23 30 30 0 30 call(s) evaluated
2026-06-07 21:21:03 30 30 0 30 call(s) evaluated
2026-06-07 21:14:47 30 30 0 30 call(s) evaluated
2026-06-07 21:08:51 30 30 0 30 call(s) evaluated
2026-06-07 21:05:33 30 30 0 30 call(s) evaluated
2026-06-07 21:00:35 30 30 0 30 call(s) evaluated
2026-06-07 20:56:20 30 30 0 30 call(s) evaluated
2026-06-07 20:53:03 30 30 0 30 call(s) evaluated
2026-06-07 20:50:00 30 30 0 30 call(s) evaluated
2026-06-07 20:50:12 30 30 0 30 call(s) evaluated
2026-06-07 20:29:07 3 3 0 3 call(s) evaluated