# Claude and OpenAI team enablement > We roll out Claude, ChatGPT, Claude Code and Codex to your teams, then teach each team to use them in its daily work, with admin setup, data rules and a library of tested workflows. - Page: https://www.ninalabs.ai/services/enablement/ - Publisher: Nina Labs AI LLC - Updated: 2026-10-08 ## Definition Team enablement turns AI licences into daily use. It covers the admin setup, the data rules, hands-on training for each role and a shared library of workflows that people trust and reuse. ## What you get - Admin setup for Claude Enterprise, ChatGPT Enterprise or both, with SSO, roles, data controls and connectors to your tools - Data rules in plain language that say what can go into which tool, and what cannot - Hands-on workshops for each role, built on that team's real tasks - A workflow library of tested prompts, Projects, Agent Skills and connectors for each team - Claude Code and Codex for engineers, with repository setup, permissions and review rules - A champions network, office hours and a 30-day adoption review with numbers ## Typical use cases - **Engineering:** Claude Code and Codex in the terminal, the IDE and CI, with CLAUDE.md and AGENTS.md files, shared Skills and clear review rules. - **Support and success:** Ticket summaries, reply drafts and knowledge-base updates in Claude or ChatGPT, connected to the help desk through approved connectors. - **Sales and marketing:** Account research, call preparation and proposal drafts from approved sources, with a person checking every claim that goes to a customer. - **Finance and operations:** Spreadsheet analysis, reconciliations and monthly reports, with the checks a controller needs before the numbers leave the team. - **Legal and compliance:** Contract review against your playbook, with the source of each finding cited and a lawyer in the loop. - **Leadership:** Briefings, research and decision memos, with clear rules on which confidential data stays out of the tools. ## Licences are not adoption Buying seats is the easy part. Use stays low when people do not know what to use the tool for, what data they may put in it, or how to check what comes out. We close those three gaps, team by team. ## Claude We set up Claude Enterprise with single sign-on, roles and data controls, and connect it to the tools each team already uses, such as drives, chat, ticketing and the code host. Teams get Projects with their own reference material and Agent Skills for the tasks they repeat every week. Engineers get Claude Code on their own repositories, with a CLAUDE.md file per repository, permission rules, hooks for your checks and a review process for agent changes. ## OpenAI We set up ChatGPT Enterprise with single sign-on, workspace roles, data controls and connectors. Each team gets shared workflows for its repeat tasks and a short guide on when to use which model. Engineers get Codex in the terminal, the IDE and the cloud, with AGENTS.md files, sandbox settings and Codex code review on pull requests. ## How a rollout runs 1. **Week 1: setup and rules.** Admin setup, connectors and the data rules, agreed with security and legal. 2. **Week 2: first team.** Hands-on workshops on the team's real tasks. We measure time per task before and after. 3. **Weeks 3 and 4: workflow library.** The workflows that worked go into a shared library, with an owner for each one. 4. **Then: the next teams.** Champions in each team, weekly office hours and a 30-day review with usage and time saved. ## When teams outgrow chat Some workflows become stable enough to run on their own. We move those into production agents on the Claude Agent SDK or the OpenAI Agents SDK, with evals, tracing and approval steps. See [AI agents in production](https://www.ninalabs.ai/services/agents/). ## Stack we use Claude Enterprise, Claude Code, Agent Skills, ChatGPT Enterprise, OpenAI Codex, Model Context Protocol (MCP), Gemini Enterprise, GitHub Copilot ## What we measure - Weekly active users by team - Workflows in the library and how often each one is used - Time saved per workflow, measured on real tasks - Data-rule incidents, which should be zero ## Frequently asked questions ### Should we choose Claude or ChatGPT? Many companies use both. The models have different strengths, and those strengths change with each release. We test both on each team's real tasks and recommend per team, with the cost of each option. ### Do you train non-technical teams? Yes. Most of our workshops are for support, sales, finance, legal and operations teams. Each session uses the team's own documents and tasks, not generic examples. ### How do you keep company data safe? We use enterprise plans, which do not train on your data by default. We set up SSO, roles and data controls, scope each connector to what the team needs, and write data rules that people can follow without a lawyer. ### How long does a rollout take? The first team is usually productive in two weeks. A company-wide rollout takes six to twelve weeks, depending on the number of teams and the tools they connect. ## Related services - [AI software factory](https://www.ninalabs.ai/services/software-factory/) - [AI agents in production](https://www.ninalabs.ai/services/agents/) ## Contact Talk to an engineer through the [contact form](https://www.ninalabs.ai/contact/) or by email at contact@ninalabs.ai.