Lead AI transformation at your company
Most companies already have someone leading AI adoption. They answer questions about which tools are approved, collect the prompts and workflows that work, share skills across teams, and get pulled into budget conversations. Usually that person is a staff engineer, an engineering manager, or an operations lead, and the work sits on top of their actual job.
That arrangement works for a while. It stops working once AI spend becomes a real line item, several teams buy overlapping tools, and leadership starts asking what the company gets in return. At that point the work needs an owner with a mandate, a budget, and a way to measure results.
Below is a template job description for that role. If you are the person doing this work informally, you can adapt it, bring it to your leadership, and use it to propose that the role become yours.
AI Director
- Department: Engineering, Operations, or Technology
- Reports to: CTO, CIO, or COO
- Location: [Location / Remote]
- Type: Full-time
Role summary
We are looking for an AI Director to lead how our company adopts and gets value from AI. This person moves us from scattered experimentation to measured, repeatable improvements in how work gets done.
This role is not about buying licenses or running generic prompt workshops. The AI Director works with leaders across the business to baseline how work happens today, redesign key workflows around AI with the right human oversight, and measure the results. They own our AI spend and are accountable for showing what it returns.
Giving everyone a license is not the same as changing how work gets done. High login rates can hide the fact that most usage is shallow, that the gains are uneven across teams, and that nobody can connect AI spend to delivery. Three problems tend to show up:
- Uneven proficiency. People have access to capable tools but use them for small tasks, so the value stays low.
- Fading gains. Early improvements drift back toward old habits when nobody is measuring them.
- Shadow AI. When approved tools are hard to use or slow to arrive, people use unapproved ones, which creates security and compliance risk.
An AI Director addresses these by treating AI as an operating change rather than a procurement exercise: pick specific workflows, measure them before and after, and scale what works.
Responsibilities
Strategy and measurement
- Set the company's AI enablement strategy and priorities, tied to business outcomes rather than usage counts.
- Establish baselines before rolling out new tools or workflows, such as cycle time, throughput, rework, and cost.
- Define clear success measures for each initiative, for example shortening pull request review time.
- Report regularly to leadership on AI adoption, spend, and impact on delivery, and separate meaningful usage from activity that looks busy but changes little.
Enablement and change management
- Build and lead a network of AI champions across teams, with clear goals tied to real workflows.
- Create role-specific training based on the work people actually do.
- Handle resistance and inertia with structured change management, and help managers make AI part of how their teams work.
Tools, spend, and operations
- Evaluate AI assistants, coding agents, and platforms, and keep the approved toolset small and well supported.
- Own the AI budget: track spend by tool, team, and person, forecast costs, and adjust allocations based on usage and delivery.
- Identify where expensive usage is not producing results, and where more budget would help people who are using it well.
- Guide the move toward agents, including how they connect to internal systems through standards such as the Model Context Protocol (MCP).
Governance and security
- Work with security and IT to find and reduce unapproved AI use by offering approved alternatives that people want to use.
- Set and maintain acceptable use policies, access controls, and logging for AI tools and agents, aligned with frameworks such as the NIST AI Risk Management Framework.
- Work with data teams so that data used by AI tools is classified, accessible, and safe.
Workflow redesign
- Partner with leaders in engineering, finance, sales, support, and operations to find repetitive, high-friction processes.
- Break those processes into steps and decide which belong to deterministic automation, which to AI, and which stay with people.
- Decide where AI can act on its own and where a person must review or approve its output.
- Turn proven workflows into reusable AI skills that any team can install, and track which skills get used and where they help.
Qualifications
- Experience leading technology adoption or transformation across multiple teams.
- Strong working knowledge of LLMs, AI coding tools, retrieval, and agent workflows, enough to evaluate vendors and work closely with engineers. Hands-on coding is helpful but not required.
- Ability to map a process, find its constraint, and redesign it.
- Comfort with budgets and business cases, including forecasting usage-based costs.
- Clear communication with executives and individual contributors, and the ability to drive alignment without direct authority.
- Background in engineering, product, operations, or consulting.
First-year success measures
- Baselines established for the workflows and teams in scope before new rollouts.
- Full visibility into AI spend by tool, team, and person, compared against delivery.
- Sustained, meaningful use of approved AI tools across target teams.
- A working champion network with owners in each major function.
- At least three workflows redesigned and in production, each with a measured improvement against its baseline.
Making the case for the role
If you are already doing this work, the strongest pitch is evidence from your own company rather than industry statistics.
Show the current state
Gather what you can: which AI tools teams pay for, what they cost, how widely they are used, and where unapproved tools appear. Fragmented spend and uneven adoption make the need concrete.
Reframe the ask
The company does not need more AI tools. It needs someone accountable for getting value from its existing AI spend.
Commit to measurement
Propose a baseline and a target for one or two workflows, such as review time or new deliveries per developer, and offer to report on them quarterly.
Address risk
Explain how you will work with security on approved tools, access, and agent governance.
Point to what you have already done
The champion network, the guides, the tool evaluations, and the questions you already answer are the first version of this role.
Measuring the work
An AI Director is only as credible as the numbers they bring to leadership. Antenna connects AI tool usage and spend with delivery data from Git and project trackers, so you can baseline teams, track AI budgets, and see whether AI is changing how work ships. If you are building the case for this role, request a demo to see how it supports the measurement side.