@@ -29,12 +29,12 @@ GitLab's unique way of working asynchronously, handbook first method, utilizatio
*** Duo Client SDK
*** Erran Carey - Principal Engineer
** Martin Wortschack - Models, Gateway & Infra
*** Gateway Services
*** AI Health + Connectivity
*** Model Selection
*** Duo Service Infra
*** Model Evaluation Infra
*** Mohamed Hamda (Acting EM)
**** Duo Service Infra
**** Model Evaluation Infra
**** Gateway Services
**** AI Health + Connectivity
**** Model Selection
** Sebastian Rehm - Duo Agent Platform
*** Annie Ying
**** Agent Developer
@@ -76,7 +76,7 @@ These are the teams (please update them if they are stale):
| [AI Core Infra](/handbook/engineering/ai/ai-core-infra/) | Abstraction Layer / AI Gateway for LLM integration to the application (GitLab Chat, Code Suggestions and other AI capabilities) |
| [AI Model Services](/handbook/engineering/ai/ai-model-services/) | Model Selection (model lifecycle, selection engine & UI) / Health & Connectivity (Duo Health Check, setup, connectivity) / Gateway Services (Prompt Registry, Events tracking, AIGW billing) |
| [Editor Extensions: VS Code](/handbook/engineering/ai/editor-extensions-vscode/) | Maintains the GitLab Workflow VS Code Extension ([maintainers](https://gitlab-org.gitlab.io/gitlab-roulette/?currentProject=gitlab-vscode-extension&mode=show&hidden=reviewer)) & [Web IDE](https://gitlab.com/gitlab-org/gitlab-web-ide) extensions and the [language server](https://gitlab.com/groups/gitlab-org/-/epics/2431). Also contributes with UX improvements for Code Suggestions within GitLab Workflow. |
The Custom Models group focuses on additional, custom models that power GitLab Duo functionality in support of our customers' unique data and use-cases.
The AI Model Services group focuses on additional, custom models that power GitLab Duo functionality in support of our customers' unique data and use-cases.
## Mission, what we own
Custom Models is the end-to-end owner of GitLab's **customer-facing model intelligence layer**: which models are available, how they're selected, the health and connectivity of the customer experience, and the gateway service surface (prompts, internal events, and AIGW billing). One team, full stack, across Customer zero, SaaS, Self-Managed, and Dedicated.
AI Model Services is the end-to-end owner of GitLab's **customer-facing model intelligence layer**: which models are available, how they're selected, the health and connectivity of the customer experience, and the gateway service surface (prompts, internal events, and AIGW billing). One team, full stack, across Customer zero, SaaS, Self-Managed, and Dedicated.
We are responsible for all backend aspects of the product categories that fall under the [Custom Models group](/handbook/product/categories/#custom-models-group) of the [AI Powered stage](/handbook/product/categories/#ai-powered-stage) of the [DevOps lifecycle](/handbook/product/categories/#devops-stages). Our product direction is on the [Category Direction, Custom Models Management](https://about.gitlab.com/direction/ai-powered/custom_models/) page, and the features we work with are listed on the [Features by Group page](/handbook/product/categories/features/#custom-models).
We are responsible for all backend aspects of the product categories that fall under the [AI Model Services group](/handbook/product/categories/#ai-model-services-group) of the [AI Powered stage](/handbook/product/categories/#ai-powered-stage) of the [DevOps lifecycle](/handbook/product/categories/#devops-stages). Our product direction is on the [Category Direction, AI Model Services Management](https://about.gitlab.com/direction/ai-powered/custom_models/) page, and the features we work with are listed on the [Features by Group page](/handbook/product/categories/features/#custom-models).
## Organisation
@@ -85,7 +85,7 @@ The following members of other functional teams are our stable counterparts:
To make boundaries clear for teams inside and outside the org, the following areas are owned by counterpart teams. When work touches these areas, loop in the listed owner.
| Area | Owner | Relationship to Custom Models |
| Area | Owner | Relationship to AI Model Services |
| --- | --- | --- |
| Subscription purchase / buying flows, plan management | Fulfillment | Counterpart for Gateway Services. We own AIGW-side billing/metering; they own purchasing and subscriptions. |
| Product analytics tooling & instrumentation | [Analytics Instrumentation](/handbook/engineering/data-engineering/analytics/analytics-instrumentation/) | Counterpart for Gateway Services. We use their tooling for internal events and tracking. |
@@ -94,20 +94,20 @@ To make boundaries clear for teams inside and outside the org, the following are
- Label Subscription: [`~group::ai model services`](https://gitlab.com/groups/gitlab-com/-/labels?subscribed=&search=group%3A%3Aai+model+services)
### Organisational Labels
Issues owned by the Custom Models group should have these labels, as appropriate:
Issues owned by the AI Model Services group should have these labels, as appropriate:
-`~"group::custom models"`
-`~"group::ai model services"`
-`~"devops::ai-powered"`
-`~"section::data science"`
-`~"Category:Model personalization"`
-`~"Category:Self-Hosted models"`
-`~"category:ai gateway services"`
-`~"category:model selection"`
-`~"category:health & connectivity"`
-`~"category:health & connectivity"`
In addition, issues should contain the relevant `~type:` and subtype labels.
## How we work
@@ -208,7 +208,7 @@ Technical ownership of the work defined by an epic is delegated to a **Tech Lead
## Team Milestone Planning Process
Custom Models follows the [Product Development Flow](/handbook/product-development/how-we-work/product-development-flow/) and [Cross Functional Prioritization](/handbook/engineering/workflow/cross-functional-prioritization/). The team uses a planning issue and boards to manage the planning process. [Planning automation](https://gitlab.com/gitlab-org/ai-powered/custom-models/custom-models/-/blob/main/doc/planning/index.md) scripts are available to make this easier.
AI Model Services follows the [Product Development Flow](/handbook/product-development/how-we-work/product-development-flow/) and [Cross Functional Prioritization](/handbook/engineering/workflow/cross-functional-prioritization/). The team uses a planning issue and boards to manage the planning process. [Planning automation](https://gitlab.com/gitlab-org/ai-powered/custom-models/custom-models/-/blob/main/doc/planning/index.md) scripts are available to make this easier.
[Planning issues](https://gitlab.com/groups/gitlab-org/-/epics/13440) for each milestone are created by the PM and used to coordinate upcoming work between the PM, EM, and stable counterparts.
@@ -224,7 +224,7 @@ During each milestone, planning is completed for the next milestone:
### Planning Issue
Each month a planning issue is created by the PM using automation and the [Custom Models Planning template](https://gitlab.com/gitlab-org/ai-powered/custom-models/custom-models/-/blob/main/.gitlab/issue_templates/milestone-planning-template.md). This is the discussion area for the planning team members (PM, EM) for a specific milestone and links to the Planning and Build Boards.
Each month a planning issue is created by the PM using automation and the [AI Model Services Planning template](https://gitlab.com/gitlab-org/ai-powered/custom-models/custom-models/-/blob/main/.gitlab/issue_templates/milestone-planning-template.md). This is the discussion area for the planning team members (PM, EM) for a specific milestone and links to the Planning and Build Boards.
### Planning Board
@@ -261,8 +261,8 @@ All workflow statuses in the [Product Development Flow](/handbook/product-develo
| Next 1-3 Milestones | `%Next 1-3 Milestones` | `~workflow::problem validation`, `~workflow::design`, `~workflow::solution validation`, `~workflow::planning breakdown`, `~workflow::ready for development` |
| Next 4-6 Milestones | `%Next 4-6 Milestones` | Same as Next 1-3 Milestones |
@@ -276,7 +276,7 @@ The [issue triage report](https://gitlab.com/gitlab-org/quality/triage-reports/-
## Customer support & Requests for Help
To better support calls with customers (existing and prospective), Custom Models provides engineers who prioritise customer support requests, so that load and knowledge are shared across the team.
To better support calls with customers (existing and prospective), AI Model Services provides engineers who prioritise customer support requests, so that load and knowledge are shared across the team.
**RFHs are owned by the functional team whose scope they fall under**, Gateway Services handles gateway RFHs, Health & Connectivity handles setup/connectivity RFHs, and so on. Each functional team assigns a **triage DRI** who either drives the RFH to resolution or explicitly hands it off with context. Support requests should be acknowledged within 24h.
@@ -304,7 +304,7 @@ It is **not** expected for engineers in support to:
## Communication
Custom Models communicates based on the following guidelines:
AI Model Services communicates based on the following guidelines:
1. Always prefer async communication over sync meetings.
1. Don't shy away from arranging a [sync call](#ad-hoc-sync-calls) when async is proving inefficient, but try to record it to share with team members.
@@ -328,7 +328,7 @@ We operate using async communication by default. There are times when a sync dis
## Blog Posts
Blog posts written by Custom Models team members:
Blog posts written by AI Model Services team members:
-[Developing GitLab Duo: How we validate and test AI models at scale](https://about.gitlab.com/blog/2024/05/09/developing-gitlab-duo-how-we-validate-and-test-ai-models-at-scale/), [@susie.bee](https://gitlab.com/susie.bee)
-[GitLab Duo Self-Hosted: Enterprise AI built for data privacy](https://about.gitlab.com/blog/2025/02/27/gitlab-duo-self-hosted-enterprise-ai-built-for-data-privacy/), [@susie.bee](https://gitlab.com/susie.bee)