0000 M · Surface
Engine-failure prediction, deployed and actually used.
At a speed the industry isn't used to.
Product lead on mtu NautIQ Foresight and mtu EnergetIQ. 17 years building products; the last four turning marine-engine data into warnings crews act on every day.
Illustrative signal · real data and thresholds stay under NDA
1200 M · Selected work
Selected work
Marine engine-failure prediction
Live failure prediction for medium/high-speed marine diesels, built on mtu NautIQ Foresight, in daily diagnostic use and rolling out to the service organisation. The flagship story.
Case study ↗NautIQ Foresight · The platform years
Four years building the platform the prediction stands on: a vessel's live data brought ashore, self-serve analytics, and the unsexy problems, like mapping, that make it real.
Case study ↗mtu EnergetIQ · From vessels to plants
The expansion into energy systems: microgrids, power plants and the standby fleets behind data centres, with the whole plant on one live single-line diagram.
Case study ↗emetriq · Deutsche Telekom Group
Machine-learning products at telco scale, from AI-driven content buying to second-precise TV-campaign measurement, plus an internal-service-to-SaaS turnaround.
Case study ↗CEWE · Lead UX
Leading a 10-person UX team on software that carried ~85% of CEWE's ~€450M revenue. Redesigns that lifted it +11% and +18%.
Case study ↗3500 M · Sediment record
Track record
Seventeen years of building products. The proof, in numbers.
-
2022 to today
mtu NautIQ Foresight / mtu EnergetIQ
Product lead since 2022 on mtu NautIQ Foresight and mtu EnergetIQ, MTU's equipment-health-management platforms for vessels and energy systems, as long-term external partner via IO-Dock (Hamburg). The core of it: live engine data pulled from vessel/plant to shore/remote monitoring centre with minimal latency, the backbone a fleet's health monitoring runs on. More recent work: offline health systems for naval defence programmes across multiple navies (under NDA), and the live failure-prediction case this site is built around.
-
2019–2021
emetriq · Deutsche Telekom Group
Machine-learning product MVPs in the Telekom group: AI-based content buying for a major streaming platform, and second-precise TV-campaign impact tracking. Turned an internal service into a SaaS with its go-to-market, and served on the management board for organizational development.
-
2015–2018
CEWE · Lead UX
Led a 10-person team across desktop, mobile and kiosk for Europe's largest photo service. The order software my team redesigned handled ~85% of CEWE's ~€450M revenue, and the redesign lifted it +11%. The photo-book assistant followed at +18%. Shipped in 34 languages through 900 retail partners, with 21M downloads.
-
2009–2014
Foundations
The craft years: web development, interaction design and project leadership across agency and freelance work.
- 17 yrs
- 10-person team
- ~85% of ~€450M revenue
- +11% / +18% uplift
- 34 languages
Built for and inside:
3200 M
About
I'm Frederic Mejia. I turn an expert's knowledge and a fleet's raw data into predictive-maintenance product that ships fast and that the people on the tools actually use. I'm fluent on both sides of the table: the engineering the methods are built in, and the uptime, cost and risk a board funds.
Specific engagements, fleets and partners are discussed privately, under NDA.
● Contact established
Meeting at SMM Hamburg?
This September I'm taking a small number of private conversations with people who own a predictive-maintenance mandate: operators, OEMs, insurers, class societies. Tell me what a day of downtime costs you; I'll tell you candidly where I can and can't move that number.
Book a private slot3800 M
Contact
Tell me what a day of downtime costs you, and I'll tell you candidly where I can and can't move that number.