Copilot(s) and AI in QA

AI in QA

QA‑relevant AI stack for target platform:

1. • SAP (Joule)
2. • Oracle Fusion (OCI Generative AI)
3. • Microsoft Dynamics 365 (Copilot)
4. • Salesforce (Einstein)

QA‑relevant AI for Pipeline:

1. • GitHub Actions - GitHub Copilot for CI (OpeenAI)
2. • GitLab CI/CD - GitLab Duo (open‑source LLMs (Llama, Mistral) + GitLab fine‑tuning)
3. • Jenkins - Optional plugins using open‑source LLMs (Llama, Mistral)
4. • Azure DevOps Pipelines - Microsoft Copilot for Azure DevOps (OpenAI GPT‑4/4o + Microsoft Ph)
5. • CircleCI - CircleCI AI (OpeenAI)
6. • AWS CodePipeline / CodeBuild - Amazon Q Developer - (Anthropic Claude, Amazon Titan, etc.)

**QA‑relevant AI work‑tracking / test‑management / issue‑tracking

1. • Azure Boards (Azure DevOps) - Microsoft Copilot for Azure DevOps (OpenAI GPT‑4/4o; Phi)
2. • Jira (Atlassian) - Atlassian Intelligence (OpenAI models (GPT‑4 family) + RAG)
3. • GitLab Issues / GitLab Plan - GitLab Duo - (open‑source LLMs (Llama, Mistral)
4. • ClickUp - ClickUp AI - ( OpenAI GPT‑4)
5. • Monday.com - Monday AI Assistant = (OpenAI GPT‑4)
6. • Asana - Asana Intelligence - (OpenAI GPT‑4)

GitGub Family:

GitHub Copilot (Inline Autocomplete)

• Code suggest as you type inside VS Code, JetBrains, or Neovim

GitHub Copilot Chat (IDE Chat Panel)

• VS Code, JetBrains

GitHub Copilot Workspace (Browser-based IDE)

• Not tied to VS Code; runs in the browser; More “agentic” than inline or Chat

GitHub Copilot in Visual Studio (not VS Code)

• Visual Studio 2022 for C#, C++, .NET.

GitHub Copilot CLI

GitHub Copilot for Pull Requests

GitHub Copilot for Docs

Microsoft 365 Copilot Family (Microsoft Copilot is built on a deep, formal, multi‑year partnership with OpenAI.)

M365 Copilot Agent

• Microsoft 365 apps (Outlook, Teams, Word, Excel, PowerPoint, Loop)

Dynamics 365 Copilot Family (Business Apps)

• Copilot for Dynamics 365 Sales
• Copilot for Dynamics 365 Customer Service
• Copilot for Dynamics 365 Finance
• Copilot for Dynamics 365 Supply Chain
• Copilot for Dynamics 365 Field Service
• Copilot for Dynamics 365 Marketing
• Copilot for Power Apps / Power Automate (Dataverse-based)

Salesforce Einstein Family (Salesforce has a formal partnership with OpenAI)

Einstein Copilot (the “main” one)

• Einstein Copilot for Sales
• Einstein Copilot for Service
• Einstein Copilot for Marketing / Marketing Cloud
• Einstein Copilot for Commerce
• Einstein Copilot for Slack (Slack AI)
• Einstein Studio (Bring Your Own Model)
• Einstein Trust Layer This is the governance layer (similar to Microsoft Graph grounding + safety).

SAP

SAP Joule (SAP’s AI Assistant)

• SAP uses a multi‑model strategy, similar to Salesforce Einstein
• SAP has a formal partnership with Microsoft to integrate: GPT‑4, GPT‑4 Turbo, GPT‑4o, Other Azure OpenAI models
• SAP supports Anthropic Claude models for: Enterprise reasoning, Document understanding, Knowledge extraction
• SAP AI Core can host: HuggingFace models, Custom enterprise models, Third‑party LLMs

IBM ( watsonx.ai → LLM hosting + Granite models )

IBM does not compete with SAP, Salesforce, Oracle, or Dynamics. Instead, IBM positions itself as:

• The AI + governance layer that works across whatever systems you already use. This means Granite models are not in conflict with SAP Joule, Salesforce Einstein, Dynamics Copilot, or Oracle Fusion AI. They are adjacent, not competing.

IBM’s core offerings:

• watsonx.ai → LLM hosting + Granite models
• watsonx.data → data lakehouse
• watsonx.governance → AI governance, risk, compliance
• watsonx Assistant → chatbot builder
• watsonx Orchestrate → workflow automation
• watsonx Code Assistant → code modernization (COBOL, Z/OS, Ansible) Is IBM trying to make Granite the “default choice”? Yes — but only in regulated industries

Oracle

Oracle Fusion Cloud “AI Assistants” - Oracle uses a multi‑model strategy, similar to Salesforce and SAP.

• Cohere models (primary partner)

1. • Cohere Command (general-purpose LLM)
2. • Cohere Embed (embeddings)
3. • Cohere Rerank

• Meta Llama models, OCI also hosts:

1. • Llama 2
2. • Llama 3

• Mistral models, OCI supports:

1. • Mistral 7B
2. • Mixtral 8x7B

• Oracle’s own models
• Oracle has internal models for:

1. • forecasting
2. • anomaly detection
3. • supply chain optimization
4. • financial risk scoring

• Customer-provided models

Oracle’s AI Assistant (Copilot-equivalent)

• Oracle Fusion ERP AI Assistant
• Oracle Fusion Sales AI Assistant
• Oracle Fusion Service AI Assistant
• Oracle Fusion HCM AI Assistant
• Oracle Fusion SCM AI Assistant