Technical Delivery Manager - Agentic AI Engineering (Manager)
for section headings: e.g., About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule. We need to include sections only when posting has real specific content. We have content for responsibilities, skills/attributes, qualifications (to qualify), ideally also have (preferred qualifications), benefits/offer (What We Offer You includes compensation and benefits). Pay: they gave salary ranges. So we need a Pay section. Schedule: not mentioned, so omit. Benefits: they mention medical/details: comprehensive compensation and benefits package, medical/dental, pension/401k, paid time off, flexible vacation policy, holidays, etc. So Benefits section. We need to preserve all substantive info, not invent. Remove boilerplate: cookie/navigation/legal/EEO/equal-opportunity/scam-warning, duplicated lines, application instructions. So we should strip lines like "Are you ready to shape your future with confidence? Apply today..." and the EEO statement at end, also the "EY | Building a better working world" paragraph, etc. Also remove "Location: Anywhere in Country" maybe we can keep as a short
metadata? The instruction: "Do NOT put Job ID / requisition numbers as headings — fold minor metadata into a short
or drop it." Location is minor metadata; we can include a short
at top. We need to output only HTML fragment, no /, no markdown fences. Structure: maybe start with
Location: Anywhere in Country
. Then sections. Section headings: Use. For responsibilities: Responsibilities
then - ... each responsibility.
We need to preserve all responsibilities bullet points. They are listed as sentences starting with verbs. We'll turn each into
- .
Similarly Skills And Attributes For Success: could be
Skills
(or Skills And Attributes). Use Skills
. Then list each bullet.
Qualifications: "To qualify you must have ..." includes bullet-like sentences separated by commas? Actually it's a paragraph with multiple requirements separated by periods. We'll split into list items: Bachelor’s or Master’s degree..., 7+ years..., technical background..., demonstrable ability..., proven track record..., working fluency..., hands-on experience..., evidence of holding..., ability to communicate...
We'll make each a - under
Requirements
or Qualifications
. The instruction says use headings like Requirements, Qualifications, Skills, Benefits, Pay, Schedule. We can choose Requirements for mandatory qualifications, and maybe Qualifications for preferred? But they said prefer few substantive sections over many thin ones. Could combine Requirements and Qualifications into one section? But we have both mandatory and preferred. Could have Requirements
for mandatory, and Preferred Qualifications
for ideally. But they said use headings like Requirements, Qualifications, Skills, Benefits, Pay, Schedule. So we can have Requirements (mandatory) and Qualifications (preferred). Or we could have Requirements covering both, but better to separate.
Skills: we already have Skills And Attributes For Success; we can put under Skills
.
Benefits: Benefits
with list items.
Pay: Pay
with paragraph describing salary ranges.
We must not include sections with filler like "commensurate with experience". We have actual ranges.
Now we need to ensure we don't repeat content. Let's extract.
First, location line: "Location: Anywhere in Country". We'll output Location: Anywhere in Country
.
Now "The opportunity" paragraph: This is description of role. Could be considered "About the role". We'll put About the role
then with that paragraph(s). The opportunity paragraph: "We are seeking a seasoned Technical Delivery Manager to drive the day-to-day execution of the engineering teams building EY’s Agentic AI platform. This role owns delivery cadence across multiple development programs: running the Scrum process and its ceremonies, keeping work flowing, and holding engineers accountable for what they committed to and when. You will be the person who knows, at any moment, what is genuinely in flight, what is genuinely blocked, and what is quietly slipping. This is deliberately not a facilitation-only role; you must understand what is actually being built (Kubernetes and GPU infrastructure, CI/CD/CV pipelines, model serving and routing, agent orchestration, data and state layers) well enough to interrogate an estimate, tell a real dependency from a convenient one, and challenge scope and sizing with evidence. The role is ideal for an engineer-turned-delivery-lead who is credible with senior engineers because they have done the work themselves, who is comfortable being the person who asks the uncomfortable question in stand-up, and who measures success in working capability shipped against committed dates rather than in ceremonies run or boards kept tidy."
We'll put that as a single
.
Now Responsibilities: list each bullet.
Let's extract responsibilities bullet points:
- Own the delivery cadence: run daily stand-ups, sprint planning, backlog refinement, sprint reviews, and retrospectives across the engineering teams, keeping every ceremony short, decision-focused, and genuinely useful to the engineers in the room.
- Drive progress against commitments: track every in-flight item through to completion, surface slippage on the day it happens rather than at sprint end, and ensure blocked work is actually unblocked rather than quietly rescheduled.
- Hold engineers accountable to deliverables and timelines: agree explicit definitions of done and dated commitments, follow them up personally, and escalate early and credibly when delivery is at risk.
- Interrogate estimates and challenge padding: apply your own technical judgement to test sizing, decomposition, and complexity claims, separate genuine engineering risk from conservative buffering, and hold estimates to what the work actually requires.
- Break down ambiguous work: translate architecture intent, phase gates, and platform capability requirements into properly sized, independently deliverable stories with unambiguous acceptance criteria.
- Own sprint and release planning across the phased platform build: sequence work so each phase puts usable capability into engineering teams’ hands, and defend entry and exit gates rather than letting them soften under schedule pressure.
- Manage cross-discipline dependencies: coordinate hand-offs between platform and infrastructure, runtime and observability, data and state, trust infrastructure, security, and the AFE delivery cells.
- Own delivery risk and issue management: maintain a live view of risks, dependencies, and the critical path, drive mitigation to closure, and make trade-offs explicit to leadership while there is still time to act on them.
- Report delivery status with evidence: velocity, throughput, cycle time, flow efficiency, escaped defects, and burn against phase gates, reported honestly and without softening, including when the news is bad.
- Protect the team’s focus: absorb interruptions, scope injection, and competing stakeholder demands so engineers get uninterrupted delivery time.
- Partner with engineering leads and architects: on sequencing, staffing, and technical direction, with a clear boundary that they own the technical decision and you own whether the work lands on time and to standard.
We'll put each as
- .
Now Skills And Attributes For Success bullet points:
- Genuine technical depth: able to read code, follow an architecture or design review, and form an independent view of how hard a piece of work actually is.
- Constructive scepticism: willing to challenge an estimate, a blocker, or a “nearly done”, and to press for specifics, without putting the team on the defensive.
- Credibility with senior engineers: respect earned through technical understanding and sound judgement rather than through process authority.
- Strong command of Agile practice across Scrum and Kanban, with the judgement to adapt the process to the work instead of running the ceremony for its own sake.
- Comfortable holding people to commitments: direct, early, and specific in difficult conversations, with escalation as a considered step rather than a first move.
- Evidence-driven: reports what the delivery data shows, not what leadership would prefer to hear.
- Strong communicator, able to move between a technical design discussion, a sprint-level trade-off, and an executive delivery update without losing precision.
- Calm and steady under pressure, keeping a team delivering when dates are tight, scope is contested, and priorities are shifting.
We'll list each as
- under
Skills
.
Now Requirements (To qualify you must have):
List items:
- Bachelor’s or Master’s degree in Computer Science, Engineering, or a related technical field, or demonstrably equivalent depth.
- 7+ years in technical delivery, engineering project or program management, or Scrum Master roles supporting software engineering teams, with direct accountability for delivery outcomes.
- A technical background: prior hands-on experience as a software or platform engineer, or demonstrably equivalent technical depth.
- Demonstrable ability to evaluate engineering estimates, work decomposition, and technical risk independently, and to challenge them constructively with evidence.
- Proven track record running Scrum or Kanban for teams of experienced engineers, owning the full set of ceremonies and the health of the backlog.
- Experience delivering complex, multi-team technical programs with hard cross-team dependencies and phased delivery gates.
- Working fluency with modern cloud-native engineering: Kubernetes and containers, CI/CD, infrastructure-as-code, distributed systems, and API design.
- Hands-on experience with engineering delivery tooling (Jira or equivalent) and Git-based workflows, including the delivery metrics they produce.
- Evidence of holding delivery teams to timelines and quality standards while retaining their trust and engagement.
- Ability to communicate delivery status, risk, and trade-offs to senior stakeholders with clarity and candour.
We'll put under Requirements
.
Now Preferred Qualifications (Ideally, you’ll also have):
- Experience delivering AI or agentic AI systems, such as LLM-backed services, orchestration frameworks, agent tooling, or retrieval and knowledge pipelines.
- Experience running delivery for platform or infrastructure teams, where the product is the substrate other engineering teams build on.
- Experience with cost-aware delivery, quota and capacity planning, or FinOps practices.
- Scrum or Agile certification (CSM, PSM, SAFe, or equivalent), treated as useful supporting evidence rather than as a substitute for technical depth.
- Experience in regulated delivery environments (financial services, tax, audit, healthcare, risk) with audit-grade evidence requirements.
- A track record of introducing flow metrics such as cycle time, throughput, and flow efficiency, and improving them measurably.
- Exposure to multi-environment delivery across cloud, on-prem, edge, client-managed, and air-gapped targets.
We'll put under Preferred Qualifications
(or Qualifications). The instruction says use headings like Qualifications. We'll use Qualifications
for preferred? But we already used Requirements for mandatory. Could rename Preferred Qualifications as Qualifications. Let's do: Qualifications
for preferred.
Now Benefits: from "What We Offer You" paragraph:
- We offer a comprehensive compensation and benefits package where you’ll be rewarded based on your performance and recognized for the value you bring to the business.
- The base salary range for this job is: New York City, Boston, and Washington DC Metro Areas, Washington State, and Southern California offices – $172,800 to $359,000; Bay Area California offices – $180,000 to $374,000; All other offices locations in the US, including Sacramento – $144,000 to $329,100.
- Individual salaries
Skills
(or Skills And Attributes). UseSkills
. Then list each bullet. Qualifications: "To qualify you must have ..." includes bullet-like sentences separated by commas? Actually it's a paragraph with multiple requirements separated by periods. We'll split into list items: Bachelor’s or Master’s degree..., 7+ years..., technical background..., demonstrable ability..., proven track record..., working fluency..., hands-on experience..., evidence of holding..., ability to communicate... We'll make each aRequirements
orQualifications
. The instruction says use headings like Requirements, Qualifications, Skills, Benefits, Pay, Schedule. We can choose Requirements for mandatory qualifications, and maybe Qualifications for preferred? But they said prefer few substantive sections over many thin ones. Could combine Requirements and Qualifications into one section? But we have both mandatory and preferred. Could haveRequirements
for mandatory, andPreferred Qualifications
for ideally. But they said use headings like Requirements, Qualifications, Skills, Benefits, Pay, Schedule. So we can have Requirements (mandatory) and Qualifications (preferred). Or we could have Requirements covering both, but better to separate. Skills: we already have Skills And Attributes For Success; we can put underSkills
. Benefits:Benefits
with list items. Pay:Pay
with paragraph describing salary ranges. We must not include sections with filler like "commensurate with experience". We have actual ranges. Now we need to ensure we don't repeat content. Let's extract. First, location line: "Location: Anywhere in Country". We'll outputLocation: Anywhere in Country
. Now "The opportunity" paragraph: This is description of role. Could be considered "About the role". We'll putAbout the role
thenwith that paragraph(s). The opportunity paragraph: "We are seeking a seasoned Technical Delivery Manager to drive the day-to-day execution of the engineering teams building EY’s Agentic AI platform. This role owns delivery cadence across multiple development programs: running the Scrum process and its ceremonies, keeping work flowing, and holding engineers accountable for what they committed to and when. You will be the person who knows, at any moment, what is genuinely in flight, what is genuinely blocked, and what is quietly slipping. This is deliberately not a facilitation-only role; you must understand what is actually being built (Kubernetes and GPU infrastructure, CI/CD/CV pipelines, model serving and routing, agent orchestration, data and state layers) well enough to interrogate an estimate, tell a real dependency from a convenient one, and challenge scope and sizing with evidence. The role is ideal for an engineer-turned-delivery-lead who is credible with senior engineers because they have done the work themselves, who is comfortable being the person who asks the uncomfortable question in stand-up, and who measures success in working capability shipped against committed dates rather than in ceremonies run or boards kept tidy." We'll put that as a single
. Now Responsibilities: list each bullet. Let's extract responsibilities bullet points: - Own the delivery cadence: run daily stand-ups, sprint planning, backlog refinement, sprint reviews, and retrospectives across the engineering teams, keeping every ceremony short, decision-focused, and genuinely useful to the engineers in the room. - Drive progress against commitments: track every in-flight item through to completion, surface slippage on the day it happens rather than at sprint end, and ensure blocked work is actually unblocked rather than quietly rescheduled. - Hold engineers accountable to deliverables and timelines: agree explicit definitions of done and dated commitments, follow them up personally, and escalate early and credibly when delivery is at risk. - Interrogate estimates and challenge padding: apply your own technical judgement to test sizing, decomposition, and complexity claims, separate genuine engineering risk from conservative buffering, and hold estimates to what the work actually requires. - Break down ambiguous work: translate architecture intent, phase gates, and platform capability requirements into properly sized, independently deliverable stories with unambiguous acceptance criteria. - Own sprint and release planning across the phased platform build: sequence work so each phase puts usable capability into engineering teams’ hands, and defend entry and exit gates rather than letting them soften under schedule pressure. - Manage cross-discipline dependencies: coordinate hand-offs between platform and infrastructure, runtime and observability, data and state, trust infrastructure, security, and the AFE delivery cells. - Own delivery risk and issue management: maintain a live view of risks, dependencies, and the critical path, drive mitigation to closure, and make trade-offs explicit to leadership while there is still time to act on them. - Report delivery status with evidence: velocity, throughput, cycle time, flow efficiency, escaped defects, and burn against phase gates, reported honestly and without softening, including when the news is bad. - Protect the team’s focus: absorb interruptions, scope injection, and competing stakeholder demands so engineers get uninterrupted delivery time. - Partner with engineering leads and architects: on sequencing, staffing, and technical direction, with a clear boundary that they own the technical decision and you own whether the work lands on time and to standard. We'll put each as