HUMINTHuman intelligence. AI powered.
AI consulting · Berlin

HUMINT

Strategy, build, & the human side.

Independent advisory and fractional product leadership for teams that need strategy, implementation, and the bridge between them.
Available for EU and remote engagements in English, German, and Dutch.

Book an intro callSee selected work
Hands-on AI systemsProduct and implementationGDPR and EU AI Act aware
About

Before I managed the tools, I used them. As an online stylist with my own customer portfolio, I knew what the job actually demanded. Then I became the Product Manager building the platform for people doing that same work.

I now work as an AI consultant and product leader, helping small businesses and founders build systems that reduce manual work without sidelining their people.

No AI for the sake of it.
Just systems that reduce manual work and free people up to do what only humans can do.

Based in Berlin. Working in Dutch (native), English (C2), and German (C1).

Not sure where to start? That is exactly what the intro call is for.

Book a free 30-minute call
What I believe

AI should make your peoplefaster,not fewer.

I build AI systems that reduce the repetitive, low-value work that slows teams down, so people can focus on judgment, relationships, and the work that moves the business. Not automation for its own sake. Systems that make your team more capable than they were before.

What I do

Services.

Engagements

Who I work with.

I work best with small businesses, founders, and small teams who want to use AI in a way that still feels human.

That includes:

  • Small and growing businesses with repetitive operational work they want to get off their plate.
  • Founders and startups building their first AI layer, without a dedicated tech team to make it happen.
  • Retail and e-commerce teams, especially in fashion and lifestyle, sitting on data they are not yet using.

If you have identified a problem and need someone who can both think it through and build something real, that is where I come in.

When it tends to fit

Usually your team has spotted an AI opportunity but has not turned it into a real build yet. You have tested a few tools, but nothing has stuck operationally, and the project is stalled somewhere between idea, implementation, and adoption.

At that point you need hands-on help, not another slide deck.

What colleagues say

In their words.

Dina is a professional, motivated, and capable Product Owner who bridges the gap between business needs and technical execution effortlessly.
Biniam Asnake KefaleEngineering Lead
Her ability to take initiative and drive projects forward with focus and reliability has made a real impact on our team.
Diana PulnarVP Product
Dina brings a rare combination of dedication, professionalism, warmth, and humor, making her not only effective but also a joy to collaborate with.
Malin FinneCCO & Former COO
Selected work

Case studies.

Outfittery

AI Powered Stylist Platform and Automation

Product Manager, AI and platform initiatives

The online stylist team was spending a significant part of every working day on coordination that should have been handled by the system. Matching customers to stylists, routing work, chasing status. I defined the problem with operations, scoped the fix with engineering, and shipped iteratively into live workflows without disrupting the team mid-shift. From there I led product for a wider set of AI systems: automated outfit curation with a structured AB testing program against human stylists, and a feedback system that unified four disconnected internal tools into one stylist facing view with AI generated summaries.

  • Reduced manual coordination workload by 80% through automated customer-stylist assignment
  • Increased throughput from 29 to 34 orders per day through AI-assisted decision support, keeping stylists in control of every customer-facing call
  • Targeted 80%+ preview acceptance rate for AI generated outfit previews across a structured menswear and womenswear testing program
  • Unified four fragmented feedback sources into a single stylist facing view, closing a gap where customer input was getting lost before reaching decision makers
+17%
efficiency (29 → 34 orders/day)
−80%
manual coordination workload
Spottr / Capstone Project

AI Member Retention for Boutique Fitness Studios

End-to-end AI consultant engagement

Structured as a real consulting engagement from day one. Identified churn as the core commercial problem for boutique fitness studios, built and validated a working MVP, and packaged the result with the documentation a buyer needs before saying yes to any AI system.

  • Working churn prediction model at 92.5% accuracy with automated weekly coach briefings via n8n
  • GDPR data protection impact assessment covering wearable and member data
  • EU AI Act compliance pack and standalone ROI model included as standard
33 days
break-even point
1,099%
projected 12-month ROI
Based on a modelled scenario: 200-member studio, €3K implementation cost
Compliance & buyer readiness

Delivered with full compliance documentation: GDPR data protection impact assessment, EU AI Act compliance pack ready for stakeholder review, and a standalone ROI model. Everything a buyer needs before saying yes.

Python·scikit-learn·LangChain·n8n·GDPR·EU AI Act
View project ↗
Pitch deck
Spottr. Smart Member Retention.

PDF, 8 slides

Download deck ↓
Also

Selected builds.

How I work

A clear engagement.

  1. Find the coordination tax

    Before anything gets built, I look at how work actually moves. Where are people doing what a process should be doing? Where is information getting lost between handoffs? Most operational drag hides in the gaps between tools and between people, not in the tools themselves.

  2. Decide what is worth fixing

    Not every inefficiency is worth automating. I help you separate the high-cost, high-frequency problems from the ones that feel annoying but do not move the needle. Honest trade-offs, no vendor allegiance, no solution looking for a problem.

  3. Build something that works in practice

    A working slice, not a slide deck. Agents, automations, pipelines, or product features, scoped to deliver value early and designed around how your team actually works, not how a demo assumes they do.

  4. Hand it over so it sticks

    Most AI projects fail at adoption, not at build. I make sure the system is usable without me: documented, compliant where it needs to be, and understood by the people who will run it. Because a system your team does not use is just expensive infrastructure.

Credentials

Background and certifications.

Experience

  • Product Manager, internal tools and platformOutfittery4 yrs

    Shipped internal AI-assisted and operational tools used daily by online stylists and operations teams.

Certifications

  • AI Consulting & IntegrationIronHack2026

    Intensive programme covering agents, RAG, ML, automation, GDPR and EU AI Act, delivered through a full client-style capstone.

  • Microsoft Certified: AI-900 Azure AI FundamentalsMicrosoft

    Foundational certification in AI workloads, responsible AI, and Azure AI services.

Contact

Let's talk.

Based in Berlin. Available for consulting, workshops, and fractional product leadership across Europe.

Currently available for Q3 engagements.

Most projects start with a scoping conversation or a one-day workshop. From there, engagements are shaped around what actually makes sense for your team and situation. Happy to talk through what that could look like.

LinkedInGitHubBlog