The brief on day one is almost always wrong about something. Find that thing first.
Not always wrong about everything. But there’s usually one assumption that won’t survive contact with users.
Open to the right brief AI, fintech & enterprise ITSM
Fanni Csincsák — Lead Product Manager & Owner
Lead Product Manager & Owner — twelve years across AI, fintech, healthtech, and enterprise IT.
I redirect messy briefs into outcomes that hold up — on enterprise platforms where the cost of getting it wrong is real.
I ground scope in real user behaviour and push back when the brief doesn’t match the evidence — and I bring the evidence.
Currently leading ServiceNow ITSM at T-Systems International. Based in Budapest, EU-eligible. Open to roles across Europe and the UK; available for US contract work via Olively.io.
02 / Three case studies
Three case studies. Three altitudes.
The redirect at demand-level, program-level, and portfolio-level. Eight more in the full work index.
A four-phase AI/AIOps integration into enterprise ITSM
The brief on day one was a single-route integration of an external AI vendor's predictive triage. What landed was three nested redirects: severity→urgency, full→phased, single→multi-route. Each came with evidence the original brief couldn't have anticipated.
Read case study~30%
faster acknowledgement on priority incidents
~20%
faster time to resolve
68→82%
first-pass story acceptance
A fintech utility, rebuilt as an AI-driven discovery product
The brief was card vault meets discount hunter. Three weeks of user research said the product was something else entirely: people didn't want to manage cards, they wanted to discover offers worth using their cards on. The redirect was utility→discovery, with transaction-data-driven offer ranking as the new core mechanic.
Read case study24,000+
users in under six months
Live
App Store and Play Store
An AI capability, embedded inside the user's workflow
AltaML had a working FILAC classifier and Compass dataset. 20+ user interviews said legal researchers weren't going to leave their existing workflow. The redirect: destination→embedded — the capability moved into a browser extension (MyJr) that lived inside the tools researchers already used.
Read case study$1M+
in follow-on funding
20+
user interviews drove the redirect
Twelve years across product, design, and delivery.
T-Systems International
Lead PM, ServiceNow ITSM
~30% faster ack on priority incidents.
Olively.io
Founder & Principal Consultant
Senior consulting: Bankmonitor + 4 client engagements across fintech, AI, SaaS.
Kibit Solutions
Senior Product Consultant, advisory
Audits & roadmaps for banks, scaleups, multinationals.
Canua
Lead Product Designer + PO Support, fintech
~50% workflow streamlining, +70% WAU.
TradieDigital
Product Manager, CRM SaaS
CRM product line acquired by competitor.
AltaML / Jurisage
Senior Product Designer, AI venture
$1M+ follow-on funding after MVP.
Caviar
Lead Product Designer, US agency
~$70M in client funding rounds.
BUTE / BME University
IT & Product Design Instructor
800+ students, 6-year course run.
Deloitte
IT Product Consultant
Public-sector energy programme.
Across these: 800+ students taught · ~$70M in client funding shipped · 1 PhD candidacy (ABD)
On the record
What stakeholders actually said.
Recommendations from people I’ve built product with. Names and companies linked — go and ask them.
I’ve worked with a lot of product people in AI and R&D. What set Fanni apart was that she treated the brief as a hypothesis, not an instruction. She went and found out how people actually behaved, came back with evidence, and argued for a different scope than the one we’d asked for — and she was right.
Csaba Molnár
Project Manager · AI Robotics
Fanni consistently delivered above expectations, on time and within budget. Her user-centric approach resulted in a product that our customers love to use. Not just delivery — she is a problem-solver at her finest.
Victoria Luu
Copy.ai
Her intelligence and analytical approach set her apart. She digs deep into behavior analysis, ensuring every product decision is well thought out and effective.
Márcia Monteiro
Interventional Systems
She didn’t just execute what we asked for — she questioned it, tied every decision back to conversion and user behaviour, and delivered on time each time.
Janis Brix
Founder · PPC Magic
The product you delivered exceeded our expectations and demonstrated a high level of expertise. Innovative approach, dedicated work.
András Zollai
Novo Nordisk
06 / Twelve years, six opinions
Things I’ll argue about.
Hot takes earned across enterprise ITSM, fintech funnels, and AI governance. None of these will fit on a t-shirt. All of them are how I work.
Stakeholder alignment isn’t a meeting. It’s a hundred small moments where you bring data, without theatrics.
The meetings are where you confirm alignment exists. The work happens in the hallway, in the 1:1, in the Slack DM.
If your story acceptance rate is below 80%, your discovery is broken. Not your engineers.
Mine was at 68%. We fixed discovery, ran BDD/Three Amigos, and got to 82%. Engineers weren’t the problem.
Most “feature debates” are actually unspoken values debates. Surface them or watch the team grind.
When two smart people argue about a feature for an hour, they’re usually arguing about something else — risk tolerance, customer empathy, KPI weight.
The question isn’t “should we use AI here?” It’s “what happens when the model is wrong 5% of the time?”
Most AI product decisions look easy until you stress-test the failure mode. NowAssist on incident triage: easy yes. NowAssist on regulatory classification: hard no, until you have a human-in-the-loop review path. The trade-off isn’t quality vs. cost — it’s quality vs. trust.
The cost of building the wrong thing dwarfs the cost of saying “wait” for two weeks. Always.
I have receipts. Cence redirect: 3 weeks of pause, 24,000 users in 6 months. The alternative was a card vault nobody wanted.
Disagree with any of these? I’d love to hear it.
12 years
Same job
different titles
different titles
08 / The person
For twelve years I’ve led products end-to-end across enterprise IT, fintech, SaaS, and AI — strategy, agile delivery, stakeholder alignment, people leadership, vendor work, regulated platforms.
Most of the work happens in the first two weeks. Figuring out whether the brief on day one is the actual problem. Earning the right to push back when it isn’t.
The metric attached at scoping is what makes the pushback possible — not aesthetic improvements, not delivery milestones. Measurable outcomes that map back to the business problem the brief was trying to solve.
More about how I work