Blog

Part 2: Scaling Translation Inference: +82% Throughput

How we improved vLLM inference throughput by 82% using AsyncLLMEngine and right-sized continuous batching

Part 1: The Bottleneck to Scale Our Translation Inference Servers

Identifying architectural bottlenecks in FastAPI + multiprocessing setup preventing efficient GPU utilization

VoicePing Started to Provide Member Use Analyzation Dashboard, Meeting Memo, Past Invoice and Receipt

VoicePing Supports Taiwanese Companies' Expansion into Japan: Startup Island TAIWAN Achieves Cross-Language Communication at Global Matching Event

Startup Island TAIWAN works to revitalize Taiwan's startup ecosystem and support market entry into Japan. VoicePing provided real-time translation services at their matching events.

Overcoming Language Barriers in International Projects: VoicePing Implementation Case Study at Tekken Construction

As international construction projects continue to increase, communication within multinational teams has become increasingly important. In this case study, we interviewed Michael and Kim from the Management Division at Tekken Construction Co., Ltd., who have implemented VoicePing for multilingual meetings in their international projects.

VoicePing Evolved: Unlocking Business Intelligence through Mind Mapping and Mobile App MP3/Voice Memo Transcription