[Research] Add 2 elite-tier projects to Inference Engines & Serving

[alvinreal]

Project: One-API\n\n### Elite Criteria Checklist (ALL Required)\n\n- [x] Elite Criteria: ALL criteria met\n - ⭐ Stars: 31,512 (threshold: 1000+)\n - 🔄 Active: 2026-01-09 (within 6 months)\n - 🏭 Production: LLM API gateway with rate limiting, quota management, and cost tracking. Production deployments across multiple providers.\n - 📚 Quality: MIT license, comprehensive docs, stable releases\n\n### Evidence of Production Usage\n- https://github.com/songquanpeng/one-api - Used as unified API gateway for managing multiple LLM providers\n\n### Why This Belongs in Elite Tier\nOne-API solves a critical production need: unifying disparate LLM provider APIs under a single OpenAI-compatible interface. It includes rate limiting, quota management, and cost tracking - essential for production deployments.\n\n### Category\nInference Engines & Serving - High-performance Serving & API Servers\n\n---\n\n## Project: OpenLLM (BentoML)\n\n### Elite Criteria Checklist (ALL Required)\n\n- [x] Elite Criteria: ALL criteria met\n - ⭐ Stars: 12,273 (threshold: 1000+)\n - 🔄 Active: 2026-04-06 (within 6 months)\n - 🏭 Production: Enterprise-grade LLM serving platform used in production environments\n - 📚 Quality: Apache 2.0 license, full documentation, regular releases\n\n### Evidence of Production Usage\n- https://github.com/bentoml/OpenLLM - Deploy and serve LLMs in cloud environments with OpenAI-compatible API\n- https://bentoml.com - Commercial platform offering managed OpenLLM deployments\n\n### Why This Belongs in Elite Tier\nOpenLLM from BentoML provides a complete production-grade solution for running open-source LLMs. It supports 50+ models with built-in streaming, batching, and auto-acceleration. The project is backed by a commercial entity (BentoML) ensuring ongoing maintenance and support.\n\n### Category\nInference Engines & Serving - High-performance Serving & API Servers\n\n---\n\n## Summary\n\nThis PR adds 2 elite-tier inference/serving projects to the Inference Engines & Serving category:\n\n1. One-API (31,512 ⭐) - LLM API management and key redistribution system\n2. OpenLLM (12,273 ⭐) - Production-grade platform for running open-source LLMs\n\nBoth projects meet all elite-tier criteria and complement the existing inference ecosystem coverage.