Senior Director, Agentic Transformation Lead

Salesforce•Indianapolis, IN

About The Position

Salesforce is moving toward a future where AI agents do the engineering work, and that change starts with how we build software for ourselves. This role leads the AI transformation of the organization that builds and runs Salesforce's internal, employee-facing technology - and in doing so, creates the playbook every CIO in our customer base will need next. Every enterprise we're selling to will hit exactly what we're hitting right now. Pilots worked. Agents are ready. The factory floor wasn't built for them. If we solve this operationally - with receipts - we own the playbook no competitor can replicate. We solve it for ourselves. Then we sell it for a decade. Most of our engineers already use AI tools every day. The job now is to turn that broad adoption into one coordinated, measurable effort: clear standards, consistent adoption, and productivity gains that are real and visible across the whole organization - running at agent speed, not human speed. This is a senior leadership role that works largely through influence. You will partner across engineering, product, and design. Your impact comes from clarity, standards, and driving outcomes - not from managing a large headcount.

Requirements

You ship, you don't just coordinate. You have a working prototype in your hands before you have a deck.
When a decision takes more than 48 hours, you shrink the question, test it, and get the data. Speed is your default operating mode. You think in agent speed, not quarterly planning cycles. When the build loop runs faster, the org's ability to learn and course-correct moves faster too.
Outcomes are the only score. Adoption and shipped value are what count. Roadmaps, status decks, and committee approvals don't impress you. Fast failure is data, not identity.
You own it. One empowered owner who makes fast decisions, works in the open, and knocks down anything that slows down learning. You make the teams around you better without needing a large team of your own.
You work in the open. Public channels, public demos. You assume someone has already worked on this problem - you find them and build with them, not around them.
You have strong product taste. You know what makes a tool genuinely intuitive and useful, and you refuse to ship things that don't look, feel, or work beautifully for the end user.
You've led at scale and built with your hands. You have 12+ years in engineering, product, or transformation leadership - and you can still sit down and build. Real technical credibility that earns the trust of senior engineers quickly, not just titles that imply it.
12+ years in engineering, product, or transformation leadership, ideally in a large technology organization.
Deep technical credibility. You have led engineering or platform teams and earn the trust of senior engineers quickly - and you've shipped real things, not just managed people who did.
Hands-on fluency with modern AI development tools: Cursor, Claude, Agentforce, Codex, or equivalent.
A clear, evidence-based view of where productivity gains actually come from - not just awareness of the category.
Demonstrated understanding of agentic architecture: model routing, harness engineering, context-setting, and the governance and trust layer that makes orchestration reliable at scale.
A track record of driving change at scale: new standards, new ways of working, and adoption that lasts.
Strong instincts for measurement. You track what matters - frequency of ideas tested, outcomes learned, adoption rates - and avoid metrics that look impressive but don't reflect real impact.
Excellent communication. You can write a clear executive brief, run a compelling public demo, and bring a room into agreement.
Comfort with ambiguity, low ego, and high ownership.

Nice To Haves

You've built, deployed, or scaled agentic AI or automation tools inside a large enterprise engineering org.
You know Salesforce products deeply, either from working here or as a power user elsewhere.
You've defined and shipped a builder progression system - shared context files, command libraries, postmortem culture - that made a team measurably faster over time.
You've worked in a model where Slack is the operating layer: async-first, public by default, documentation-as-you-go.
You bring an existing network in enterprise engineering, platform, or AI transformation communities.

Responsibilities

Coordinate the AI effort. Bring the many separate AI and engineering initiatives across teams into one coordinated program. Reduce duplication, scale what works, and give the work a single, clear direction.
Standards and quality. Own the standards for how teams build with AI, and the process for improving them. Define what good looks like, how work is reviewed, and how expectations rise as the tools mature. Refuse to accept output that doesn't work beautifully for the end user.
Tools and adoption. Choose the set of AI development tools the organization uses - Cursor, Claude, Agentforce, and what comes next - and define the level of adoption every team is expected to reach. Keep the organization current with the best tools available.
Delivery and release. Remove the deployment, testing, and process bottlenecks that hold back the productivity AI makes possible. Modernize the path from idea to production so teams can ship in weeks, not quarters.
Measurement. Measure how work moves from idea to production and how the tools are used. Build clear, role-specific measures so impact is comparable across the org. Track what matters - frequency of ideas tested, outcomes learned, adoption - and avoid metrics that look impressive but don't reflect real impact.
Builder progression. Define and drive the system that makes every build compoundingly faster than the last: shared context files, command libraries, three-model review workflows, and postmortem culture. The builders who win aren't the ones who use AI - they're the ones who've built the system around AI.
Roles and ways of working. Define how engineering, product, and design roles change as agents take on more of the work, and shape the broader way the organization operates around it - including how we move toward a headless model where agents run the work and humans make the judgment calls.