Evaluating the options for simplifying the "AI impact" analytics implementation for self-management.

Problem

The new AI Impact is dependent on ClickHouse.

For .com customers ClickHouse is available seamless, but self-manage customers need to bring their own CH for these features.

There are several problems with BYO CH:

1. The implementation and maintenance is complex, resulting high TCO for customers.
2. It extends the adoption time and decreases Gitlab's competitiveness.
3. Its create a fragmented architecture that lacks a unified approach, making it difficult to support, adding risks for breaking change and adds barriers for innovation.

Optional solutions:

1. Proposal 1 - Cloud Connector for Optimize group analytics.
2. Proposal 2 - Implementation based on existing GL tooling (Redis, PosgreSQL)
3. Proposal 3 - Adding CH instance as a new service behinds the AI gateway
4. Today- BYO clickhouse
5. ❓

Evaluation Matrix

Factor (1 is low, 5 is high)	Weight	1) Cloud Connector for clickhouse	2) Redis, PosgreSQL	3) Service behinds the AI gateway	BYO clickhouse (today)	Sum
Impact
Reach
Effort		4	4		2
Time to market		4	4		2 (some tweaks might be needed)
Scalability		5 (we're responsible for it, instance sizing, etc.)	3 (scales with PG, won't work well for large instances)		1 (we don't manage it)
Security		4 (red data, CH Cloud is not FEDRAMP certified, 3rd party vendor)	1 (no new component, we use PG)		2
Ease of Implementation		3	4 (PG BG aggregation)		2 (mostly done)
Customizability
Total score

This page may contain information related to upcoming products, features and functionality. It is important to note that the information presented is for informational purposes only, so please do not rely on the information for purchasing or planning purposes. Just like with all projects, the items mentioned on the page are subject to change or delay, and the development, release, and timing of any products, features, or functionality remain at the sole discretion of GitLab Inc.