Add 16 missing projects with active Red Hat contributions

[soyr-redhat]

Summary

Adds 16 projects with verified active Red Hat employee contributions to data/projects.json, and introduces a new AI/ML category to data/categories.json.

Red Hat contributor presence was verified by checking @redhat.com commit email addresses in each repository's recent history.

New AI/ML category projects

• vLLM — high-throughput LLM inference engine (6/100 recent commits from Red Hat)
• llm-compressor — LLM compression library for vLLM (Neural Magic, now part of Red Hat)
• Open Data Hub — AI/ML platform, upstream of OpenShift AI (35/100)
• KServe — distributed AI inference on Kubernetes (46/100)
• TrustyAI — AI explainability and fairness toolkit (22/100)
• Model Registry — MLOps model metadata management (6/100)
• Caikit — AI model runtime toolkit (3/100)

New Development category projects

• Tekton — cloud-native CI/CD pipelines (33/100)
• Operator Framework — Kubernetes operator toolkit (40/100)
• Konveyor — application modernization for Kubernetes (75/100)
• crun — OCI container runtime (82/100)
• Konflux — trusted software build factory (34/50 in build-service repo)

New Platform category projects

• MicroShift — edge-optimized OpenShift (72/100)
• bootc — image-based Linux boot and upgrades (21/100)
• rpm-ostree — hybrid image/package system (40/100)

New Operations category project

• Open Cluster Management — multi-cluster Kubernetes management (44/100)

Each project is added as an individual commit for easy review.