Service 04
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.
In one paragraph
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.
Deliverables
What you get
- 01 Admin setup for Claude Enterprise, ChatGPT Enterprise or both, with SSO, roles, data controls and connectors to your tools
- 02 Data rules in plain language that say what can go into which tool, and what cannot
- 03 Hands-on workshops for each role, built on that team's real tasks
- 04 A workflow library of tested prompts, Projects, Agent Skills and connectors for each team
- 05 Claude Code and Codex for engineers, with repository setup, permissions and review rules
- 06 A champions network, office hours and a 30-day adoption review with numbers
Where it fits
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.
How we work
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
- Week 1: setup and rules. Admin setup, connectors and the data rules, agreed with security and legal.
- Week 2: first team. Hands-on workshops on the team’s real tasks. We measure time per task before and after.
- Weeks 3 and 4: workflow library. The workflows that worked go into a shared library, with an owner for each one.
- 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.
FAQ
Questions about claude and openai enablement
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
Often combined with
-
Service 03
AI software factory
We set up coding agents across your delivery pipeline, from spec to merge. Your engineers own the specs, the review gates and the releases.
-
Service 01
AI agents in production
We design, build and run AI agents that work inside your systems. They read, decide, call tools and hand the case to a person when a rule says so.
Tell us which process you want to hand to an agent
A 30-minute call with an engineer. We will tell you whether an AI system is the right tool for it, and what it would take to run it in production.