Product ownership · applied AI · operational systems

Jonathan Thangadurai Selvaraj

Technical Product Owner

My experience covers AI product ownership, applied-research programmes, organisation strategy and enterprise integration. I work across discovery, technical delivery, stakeholder coordination, validation and production support.

Amsterdam, NetherlandsTechnical Product Owner · AI, operations & systems

Explore my experience View my CV ↗

At TNO, I led an AI market-intelligence product from discovery through beta and handover and evaluated an eight-year AI programme to inform reinvestment and successor-programme priorities. Before that, I designed and supported enterprise integrations for manufacturing operations at Flex.

Jonathan overlooking a historic courtyard

Experience

Professional experience

TNO
Netherlands Organisation for Applied Scientific Research
The Hague, Netherlands · September 2024–August 2026
At TNO, I worked across organisation strategy, multi-partner programme delivery, AI programme evaluation, product ownership and additional advisory work with research, government, industry and business teams.
Product Owner — Market Intelligence PlatformJanuary 2026–August 2026
- Led 20+ discovery and prioritisation workshops across 10 stakeholder groups, including six external companies, to map workflows, test assumptions and rank use cases by decision value, feasibility and readiness.
- Owned the roadmap and backlog for a five-person AI team, translating discovery into user stories and acceptance criteria while running fortnightly stakeholder alignment and four-week validation and UAT cycles.
- Designed a human-in-the-loop operating model that linked LLM outputs to source evidence, tested extraction against ground truth and fed analyst review back into validation logic and roadmap decisions.
- Narrowed 15 GenAI experiments to seven beta capabilities and three priority investments; moved the platform into beta with approximately 25 participants and handed over the roadmap and feedback cycle.
- Built a value model covering review effort, processing volume, analyst coverage and refresh speed; early estimates indicated 60–85% less first-pass review time and two to four times greater company coverage.
AI Strategy & Programme Evaluation2026
- Evaluated an eight-year organisational AI programme by reviewing collaboration outcomes, programme evidence and stakeholder perspectives.
- Analysed lasting outcomes, failure modes, sector conversion and funding leverage to inform reinvestment and successor-programme priorities.
Additional advisory work — Data maturity2026
- Conducted data-maturity assessments with Data & Analytics, identifying data management, stewardship and quality constraints affecting the organisation's ability to scale AI.
Project Manager — Air Quality & EmissionsMay 2025–December 2025
- Coordinated five applied-research projects involving customers, government organisations, industry partners and internal specialists.
- Negotiated a reduced scope after a consortium partner withdrew, preserving customer value and securing an outcome accepted by the customer, partner and research team.
- Scoped the transition of an HPC-dependent air-quality model into a smartphone-accessible service for non-technical users in Kenya.
Organisation Strategy Consultant — Geological Survey of the NetherlandsSeptember 2024–April 2025
- Mapped the value chain, products, data sources, system dependencies and accountable owners across six departments with Enterprise Architecture.
- Used A3 and 5 Whys workshops to turn recurring handover, ownership and prioritisation problems into a repeatable governance framework and shared Confluence catalogue.
- Embedded the framework in cross-department planning and trained six deputy managers to take implementation forward.
Flex Ltd.
Integration Solutions Consultant
India · Southeast Asia · Mexico · October 2020–July 2022
This is where I learned what dependable technology means under real operational pressure: designing, releasing and supporting the data flows behind factory, quality, warehouse, workforce and supply-chain operations.
Integration Solutions ConsultantOctober 2020–July 2022
- Designed and delivered 10+ MuleSoft integrations connecting PLM, MES, QMS, WMS, ERP, workforce and identity systems across international manufacturing operations.
- Translated operational needs into process maps, data mappings and interface specifications, choosing API, batch or event-driven patterns and coordinating releases across four controlled environments.
- Owned production support and incident investigation using centralised logs, SQL, Postman and ELK, working directly with factory and application teams when live flows failed.
- Coordinated a move from on-premise servers and databases to a Zoho cloud platform, working with the vendor to close capability gaps and shape the expansion roadmap.
- Reduced integration processing time by 70% and eliminated recurring skipped batch records by redesigning selection logic around reliable cloud-ingestion timestamps.

Work

Selected products and projects

01Beta delivered · handed over

AI product ownership · Market intelligence

Market Intelligence Platform

From fragmented information and experiments to a governed beta product

I identified fragmented market and customer intelligence across spreadsheets and SharePoint, then defined a shared platform for company comparison, market-pull analysis and portfolio-whitespace discovery.

I led 20+ workshops across 10 stakeholder groups, owned the roadmap and backlog for a five-person AI team, and designed the human-in-the-loop evidence layer: schemas, prompts, classification logic, source traceability, ground-truth testing and validation rules.

I narrowed 15 experiments to seven beta capabilities and three priority investments, moved the product into beta with approximately 25 participants, and handed the roadmap and feedback cycle to Market Strategy.

Fortnightly alignment and four-week UAT cycles
Discovery included six external companies
Early model estimated 60–85% less first-pass review time
Early model estimated 2–4× greater company coverage

02Industry capstone

Data spaces · Circular economy

MineMatch

Exploring how mining waste can become valuable input for another organisation

For a Solita industry capstone, I led a five-person international team developing a data-space concept for matching mining-sector material and waste streams with potential users.

I combined project leadership and stakeholder negotiation with hands-on platform work, designing APIs and traceable data-sharing flows and prototyping Digital Product Passports on Hyperledger Fabric.

The project tested both the technical platform and the commercial, sustainability and trust conditions required for cross-organisational adoption.

Five-person international team
Node.js, TypeScript and PostgreSQL prototype
Multiple mining-sector data providers
Aligned with EU sustainability goals

Other projects

03Developer

IoT · Cold-chain logistics

VaxiChain

A proof of concept for safer vaccine transport, combining simulated sensor data, MQTT and OpenHAB to monitor cold-chain conditions and trigger alerts when temperatures moved outside the safe range.

Kotlin · Mosquitto MQTT · Telegraf · InfluxDB · Grafana · OpenHAB · Docker

04Project manager

IoT · Public infrastructure

Dart & Charge

An RFID and IoT-based toll-collection prototype designed to reduce waiting time and congestion at toll plazas. I led the project from concept through a working national-level demonstration.

Winner, Smart India Hackathon · #1 of 150+ teams · prototype-development grant

05Developer

Renewable energy · CINI Challenge, Italy

Beehive

A community renewable-energy concept exploring how local production and consumption data could support more coordinated energy distribution, developed as a proof of concept for the CINI Challenge.

React · Kotlin · Python · international challenge project

About

Background and approach

I started in enterprise integration at Flex, designing and supporting data flows across manufacturing, quality, warehouse, workforce and business systems.

At TNO, my work expanded into organisation strategy, multi-partner project delivery, AI programme evaluation and AI product ownership. My responsibilities included discovery, roadmaps, backlogs, validation, governance, programme analysis and product handover.

I often work across teams with different requirements. I connect user needs, technical constraints, operational processes and ownership so that decisions and delivery responsibilities are clear.

Areas I consider in product decisions

Technology: Architecture, data, interfaces, constraints and operational reliability.
People: Users, teams, incentives, decision-making and the work surrounding the product.
Institutions: Ownership, governance, policy and collaboration across organisational boundaries.
Adoption: Demand, sponsorship, evidence and the path from a promising prototype to dependable use.

Education

Education and recognition

Fully funded Erasmus Mundus scholar

2022–2024

Erasmus Mundus Joint Master's Degree — SE4GD

MSc Software Engineering & Sustainability

Selected for a fully funded Erasmus Mundus scholarship covering tuition and living support for an international programme focused on software engineering, sustainability and the European Green Deal.

Vrije Universiteit Amsterdam
LUT University
University of L'Aquila

2016–2020

Anna University

Bachelor of Technology in Information Technology · graduated Cum Laude, top 10%

My foundation in software engineering, systems development and applied technology.

Languages

English: C2
Dutch: A2 · studying toward B1
Tamil: C2
Hindi: A2

Selected recognition

Smart India Hackathon winner: First of 150+ teams · RFID toll prototype · 2019
Tamil Nadu Government grant: Early blockchain applications for public services · 2018
Solve the SDGs runner-up: SpinGreen circular-fashion concept with Spinnova · 2023

Capabilities

Capabilities and tools

Product discovery and direction

Problem framing, user and customer discovery, value-based prioritisation, roadmaps, backlogs, user stories, acceptance criteria and UAT.

AI and data products

Human-in-the-loop workflows, schemas, prompts, classification, ground-truth evaluation, validation, provenance and source traceability.

Operational systems

Enterprise integration, system boundaries, data flows, release coordination, production support and reliability across factory and business platforms.

Transformation and adoption

Operating models, governance, root-cause analysis, resistance management and alignment across business, research, government and industry.

Tools and methods

The tools, platforms and delivery methods I have worked with.

Product & delivery: Discovery · roadmaps · backlogs · user stories · acceptance criteria · UAT · Agile/Scrum · value models
AI & data: Human-in-the-loop workflows · LLM evaluation · ground truth · schemas · prompts · Python · SQL · Pydantic · PostgreSQL
Integration & operations: MuleSoft · REST/SOAP · DataWeave · batch · events · RabbitMQ · Azure Event Hubs · MES · QMS · WMS · ERP · IAM
Analysis & transformation: Service blueprints · user journeys · BPMN · ArchiMate · Obeya · A3 · 5 Whys · operating models · governance
Platforms & tools: Azure · Azure DevOps · Jira · Confluence · draw.io · Postman · ELK · Zoho · Docker · MQTT · Hyperledger Fabric

Contact

I am always interested in thoughtful conversations about product creation, applied AI, research-to-product work, public innovation and complex systems.

Write to me Read my CV ↗

Email: jonathanthangadurai@gmail.com
Telephone: +31 6 3057 3697
LinkedIn: /in/jonathanthangadurai ↗
GitHub: /JonathanThangadurai ↗
CV: Download PDF
Location: Amsterdam, Netherlands