Preterview
PulseAugur coverage of Preterview — every cluster mentioning Preterview across labs, papers, and developer communities, ranked by signal.
4 day(s) with sentiment data
-
Silero VAD endpointing issues cut off 23% of voice app answers
A voice application using Silero VAD experienced significant issues where its endpointing mechanism prematurely cut off user responses. The default 700ms silence detection threshold was too short for users who paused to…
-
Claude prompt caching fails due to timestamp, tool ordering bugs
A developer encountered a significant issue with Claude's prompt caching feature, resulting in a 0% cache hit rate across 1,412 calls. The problem stemmed from three distinct bugs: a dynamic timestamp in the system prom…
-
LLM-as-a-judge systems default to high scores; checklist approach improves accuracy
A user discovered that their LLM-as-a-judge system, using Claude, was consistently assigning scores of 7 or 8 out of 10 to answers, with 68% of 1,140 responses falling into this range. This indicated the model was not a…
-
Claude prompt caching fails due to timestamp bug, costing users more
A developer discovered that Claude's prompt caching feature was not working as intended, resulting in a 0% hit rate and increased costs. The issue stemmed from a dynamic timestamp included in the system prompt, which ma…
-
Voice AI latency slashed from 4.2s to 780ms via streaming and caching
A developer significantly reduced voice AI latency from over 4 seconds to under 1 second by optimizing various components of the system. Key improvements included streaming the first sentence of the LLM's response direc…