Technical Product Manager, Agentic AI Security

Technical Product Manager, Agentic AI Security

• The Dell Security & Resiliency organization manages the security risk across all
• aspects of Dell’s business. We are currently experiencing incredible growth to
• meet the security needs of the world’s largest technology company. With team
• members located in over 15 countries, you will have an excellent opportunity to
• influence the security culture at Dell and further develop your career.

• Join us as a Technical Product Manager, Agentic AI on our Dell Security &
• Resiliency Team in Round Rock, TX to do the best work of your career and make a

profound social impact

What you'll achieve:

• We are seeking a Principal Technical Product Manager to hold the specification
• for security workflows being delivered as AI-native capabilities. This role owns
• what gets built and why: the requirements, the measurable outcomes, the
• acceptance criteria, the scope, and the standard a workflow must meet before it
• is ready to move into design and build.

• The specification is the contract between the people whose process is changing
• and the engineers changing it. This role works with domain functional experts to
• establish what the work is, how it is performed today, and what a correct
• outcome looks like, and works alongside engineering to keep the specification
• buildable as design proceeds.

• Specifications are expected to be written, verifiable, and current. A
• requirement that cannot be verified, an outcome that cannot be measured, or a
• scope boundary that has not been stated is not complete.

Objectives of This Role

• A written specification for every workflow entering delivery, stating the

requirements, the scope, and the outcomes the capability must produce

• Acceptance criteria that state what the capability must do, expressed so that

meeting them can be verified rather than debated

• Measurable outcomes agreed with the functional owner before work starts, and

evidenced after release

• Scope held stable through delivery, with changes made deliberately and their

cost stated

• You will:
• Own the specification for each workflow, covering the problem being solved,

the requirements, the scope boundaries, the constraints, and the outcomes the

capability must produce

• Work with domain functional experts to establish how the work is performed

• today, what the process exists to achieve, where it varies, and what a
• correct result looks like, including the practice that has never been written

down

• Define acceptance criteria that state what the capability must do to be

accepted, expressed in terms that can be verified rather than interpreted

• Define the measurable outcomes for each workflow, including the baseline they

are measured against and the improvement expected

• Own the definition of ready: the standard a specification must meet before a

workflow moves into design and build, and hold specifications to that

standard rather than allowing incomplete work forward

• Scope the work: separate what is required from what is desirable, state

explicitly what is out of scope, and decompose large workflows into

increments that can be delivered and used

• Resolve conflicting requirements between functions rather than accumulating

them into the specification, and record the decision and the reasoning behind

it

• Keep the specification current as design and delivery proceed, managing

change deliberately and stating the cost of each change rather than absorbing

it

• Work alongside engineering through design and build so that specifications

remain buildable, and revise the specification when a technical constraint

makes the original requirement impractical

• Validate delivered capability against the specification with the functional

owner, and confirm after release whether the stated outcomes were achieved

• Maintain the specification standard, templates, and reference examples so

that each workflow after the first takes less effort to specify

Take the first step toward your dream career

Every Dell Technologies team member brings something unique to the table.

Here's what we're looking for in this role

• Essential Requirements
• 8+ years in technical product management, product ownership, or business

analysis for software systems, with specifications that engineering teams

built directly from

• Demonstrated ability to write requirements and acceptance criteria that are

unambiguous and verifiable, and that do not require the author present to

interpret them

• Experience eliciting requirements from domain experts, including surfacing

the tacit practice that experienced people apply without documenting

• Experience defining measurable outcomes, including establishing a baseline

and agreeing in advance what improvement will be measured and how

• Experience scoping and decomposing large bodies of work into deliverable

increments, and holding scope through delivery under pressure to expand it

• Technical fluency sufficient to work with engineers on feasibility,

constraints, and trade-offs, and to understand the systems being specified

• Working knowledge of AI-native systems, including why their behavior is

probabilistic and what that means for how requirements and acceptance

criteria must be written

• Demonstrated ability to resolve conflicting requirements between stakeholders

and hold that decision

• Written communication of a standard that engineering can build from without

repeated clarification

• Bachelor’s degree in computer science, engineering, or a related field, or

equivalent practical experience

• Desired Requirements
• Experience in product management or business analysis within a security

domain

• Experience specifying AI or machine learning capabilities for production use
• Experience with process analysis or process re-engineering, including

documenting current state and defining a target state

• Prior engineering or technical delivery background
• Experience establishing specification standards or templates where none

existed

• Familiarity with LLM behavior, agentic systems, or AI evaluation practice