Automation-first growth analytics · Berlin

I automate the boring half of analytics — and find the growth in the rest.

I'm Nonso Ezeoma, a Business Data Analyst who turns data into decisions that move the numbers that matter — conversion, engagement, retention — and builds the pipelines and AI workflows that keep them running. I don't stop at the finding; I surface the value it unlocks and the action it points to. Most analysts do one half. I do both.

See the work Get in touch

Nonso Ezeoma reviewing analytics and AI-automation dashboards

254K

user sessions analysed for conversion

automations & dashboards built end-to-end

LLMs wired into production workflows

2×/day

pipelines running unattended

Selected work

Three analytics deep-dives and two production automations — each one live and interactive, with the code behind it. Together they cover the whole funnel: acquisition, conversion, engagement, retention, pricing, and the automation that removes the manual work around it.

Growth analytics

Trivago — conversion analysis

Analysed 254,177 hotel-search sessions to find where the funnel leaks. Engagement was rising while bookings fell — I pinpointed the drop at the booking step and the segments driving it (including a high-volume device converting at just 0.65%).

+1,270more bookings from a 0.5-pt lift — traffic they already pay for

Live dashboard Code

Product analytics

Mobile — engagement & retention

Analysed 700 users to answer what separates a light user from a power user. The finding: engagement is a behaviour, not a demographic — apps installed correlates 0.98 with engagement, while age correlates 0.00. The lever is adoption, not audience.

0.98the lever that grows engagement — invest in adoption, not audience

Live dashboard Code

Geographic & pricing

NYC — rental market analysis

Read the market for a real-estate client across 17,614 listings, for renters and owners both. Two separate models showed price is set by location and room type — but bookings aren't: price and occupancy correlate −0.03. The lever owners reach for isn't the one that works.

−0.03proof discounting won't fill calendars — so owners stop cutting price

Live dashboard Code

Automation · Node.js

Job-search robot

A Node.js service that scans four job-board APIs and my inbox daily, scores every opening against a profile with the Claude API, tailors a CV for strong matches, and logs it all to a tracker — running twice a day, unattended. Built and debugged end to end.

~400jobs scored per run — a production service I built and run daily

Live dashboard Code

Automation · n8n

Lead qualification & routing

An n8n workflow that reads every incoming sales lead, scores it hot/warm/cold with an LLM plus a drafted reply, and routes the hot ones straight to Slack. Manual screening that took hours now happens in seconds, with 100% of leads reviewed by one consistent rule.

hours → secondsfor a hot lead to reach sales — a live workflow, not a demo

Live dashboard Code

01·b

More projects

Foundational analytics work from my training — across Excel, Tableau, Python and SQL. Each opens the full project write-up.

Excel Video-game sales — GameCo View project Tableau Influenza season forecasting View project Python Instacart customer behaviour View project SQL Rockbuster movie-rental analysis View project

How I work

Most analysts request a pipeline and wait. I build it. The result is analysis that ships as something a team can actually use — and keeps running without me.

Automate the manual half

Node.js, n8n and LLM APIs to kill the copy-paste — data pulls, scoring, reporting that runs itself on a schedule.

Find the growth story

SQL and Python to dig past the dashboard into the real driver of conversion, engagement and retention — the non-obvious one.

Ship it interactive

Findings people can click through and act on — not a PDF nobody opens. Framed in the one thing that matters: value.

SQLPython · pandasPower BITableauLookerNode.jsn8nClaude / LLM APIsGoogle APIs

About

I'm a Business Data Analyst based in Berlin, with a degree in Business Information Systems — a background that sits right on the bridge between the business question and the technical build.

I've worked across both B2C and B2B, and earlier operational and people-management experience shapes how I work: I care less about the chart and more about the decision it changes and the value that unlocks. What I enjoy most is removing the repetitive, manual half of analytics so growth and product teams can act on conversion, engagement and retention faster — and building the automation that keeps it running.

English at a native level, German at B2. Open to Data, Product and Marketing Analyst roles in Berlin, remote in Germany, or relocation.

Let's talk about your funnel.

If your team is drowning in manual reporting or sitting on data it hasn't turned into decisions, I'd like to help. Open to roles and interesting problems.

Email me GitHub LinkedIn