End-to-end delivery
I connect discovery, planning, execution, verification, release, and follow-through so software projects move as one delivery system.
Product & Delivery Manager · Agile Coach at NTT DATA
I help people and organisations deliver better software.
At NTT DATA, I lead and support software projects end to end and help organisations build more systematic delivery practices. Through personal projects, I explore how AI can reduce repetitive work and remove obstacles for the people producing value, while judgment, ownership, review, approval, and release authority remain human-led. I turn delivery experience into practical learning, workshops, and training.
How I work
Product & Delivery Manager · Agile Coach at NTT DATA
I lead and support end-to-end software project delivery across complex enterprise environments. I help organisations build more systematic delivery practices through clear ownership, disciplined planning, cross-functional alignment, technical awareness, and continuous learning.
People and technology
I stay close to both the people doing the work and the systems being built, connecting delivery decisions with technical reality.
Controlled change
I prefer small, verifiable changes with clear ownership, release boundaries, and a practical path back when something fails.
Practical learning
I use training, workshops, and day-to-day delivery work to help teams examine ideas and test what is useful in their own context.
AI-assisted software delivery
I explore AI-assisted delivery through real projects.
My personal, non-commercial projects are a practical environment for exploring how AI can reduce repetitive work, remove obstacles to productive work, and create useful learning for people and organisations. Codex supports discovery, implementation, testing, release preparation, and operational follow-through; judgment, ownership, review, approval, and release authority remain with people.
Scrum Assessments App
I am building an offline-first Scrum study app that separates deliberate Practice from timed Exam Simulation, then connects review back to original explanations and the Scrum Guide.
- AI-assisted work
- Implementation, testing, content validation, release documentation, and evidence collection.
- Human controls
- Product judgment, source review, device checks, subscription validation, App Store submission, and release approval remain human-led.
- Verified outcome
- Published on the App Store for iOS and Mac.
Musiki Cemiyeti (opens in a new tab)
I develop Musiki Cemiyeti as a human-directed publishing system for an independent music publication. Editors set the direction and brief; controlled automation carries out repeated editorial operations within that direction across WordPress and Instagram, with safeguards around every release.
- AI-assisted work
- Discovery, implementation, workflow design, testing, release preparation, and operational follow-through for controlled editorial automation.
- Human controls
- Editors define the brief and make editorial and publishing decisions. Automation operates inside that direction; approvals, release authority, and production checks remain human-led.
- Verified outcome
- A live self-hosted publication with controlled WordPress and Instagram operations, release safeguards, and a working human-approval path.
Simple Fasting App
I own and develop Simple Fasting App as an independent product built around a minimal experience: download it, open it, and start fasting, without unnecessary subscription pressure or a lengthy onboarding flow.
- AI-assisted work
- Codex supports coding, testing, security review, release preparation, App Store imagery, and the publishing workflow.
- Human controls
- Product decisions, code and security review, real-device verification, compliance checks, App Store submission, and release approval remain human-led.
- Verified outcome
- A live App Store product with real-device QA, widgets, Live Activities, and a controlled release record.
PasifAgresif.com (opens in a new tab)
I approached the recovery and modernization of this long-running legacy WordPress publication as a controlled delivery problem: manage migration complexity, compatibility and security risk, then move to a modern, standards-compliant implementation without losing the site’s intended visual character.
- AI-assisted work
- System investigation, migration planning, compatibility work, testing, security review, and release preparation.
- Human controls
- Scope, compatibility trade-offs, credentialed operations, production approval, rollback readiness, and live checks remain human-led.
- Verified outcome
- A live modernized WordPress publication with its intended visual character preserved and controlled migration evidence retained.
VengefulGhoul.com (opens in a new tab)
I restored a previously non-functioning WordPress site to a working state through a controlled recovery, with compatibility checks and live verification before treating the work as complete.
- AI-assisted work
- System investigation, recovery planning, implementation support, compatibility testing, and release preparation.
- Human controls
- Recovery scope, credentialed operations, compatibility acceptance, production approval, and final live checks remain human-led.
- Verified outcome
- A previously non-functioning site restored to a working live state with compatibility and recovery checks completed.
Production System
I treat the environment around software work as production infrastructure: isolated, repeatable development environments; explicit testing, security, and verification practices; and task-focused AI, subagent, and plugin-assisted workflows.
- AI-assisted work
- Task-focused agents and plugins support investigation, implementation, testing, and evidence collection inside isolated project environments.
- Human controls
- People define scope, permissions, security requirements, acceptance criteria, and release authority; automation works inside those boundaries.
- Verified outcome
- A repeatable working practice that separates projects, makes verification explicit, and keeps production actions approval-led.
Product Delivery Operating System
I use Notion as the source of record for work, living knowledge, rules, and decisions. Codex supports task execution and automation inside that human-defined structure, connecting delivery activity with the context and governance around it.
- AI-assisted work
- Codex supports scoped task execution, evidence capture, follow-through, and automation within the recorded plan and rules.
- Human controls
- People define the rules, governance, priorities, decisions, approvals, and exceptions. Automation operates inside that structure.
- Verified outcome
- An active delivery record connecting work, living knowledge, decisions, evidence, and human-defined operating rules.
These projects turn AI-assisted delivery into observable practice. I use the lessons to help people and organisations examine where AI can support productivity without weakening human judgment, ownership, review, approval, or release control.
Discuss AI-assisted delivery110participants across Anadolu Group companies and NTT DATA teams
Education
I design, deliver, and facilitate practical learning.
I turn delivery work into practical learning, workshops, and training on Agile software project management and Product Ownership. I design the learning experience, deliver the material, and facilitate participants as they apply ideas to real work. Across Anadolu Group companies and NTT DATA teams, I have worked with 110 participants in one-, two-, and five-day formats.
Credentials and continuous learning
Learning made visible through practice and certification.
My credentials bring together Scrum, product ownership, Kanban system design, and focused OpenAI partner training. The full record distinguishes professional certifications from completed training.
PSM II
SAFe 6 POPM
Kanban System Design- OpenAIOpenAI · 8 trainings
Get in touch
Let’s explore what we could develop together.
If there is a solution to develop or an area where my delivery, AI-assisted software, or education experience could help, share the context and let’s start a conversation.
Start a conversation