Hamburg · Digital Business & Data Science

Open to Werkstudent & internship roles · 12 months · Hamburg

Data & AI
analytics.

Python, SQL, Power BI and PostgreSQL, with governed AI on top. Looking for a Werkstudent or internship role.

GitHub ↗

01 — About

Hi, I'm Vansh.

Fourth-semester B.Sc. Digital Business & Data Science student combining Python, SQL, Power BI and PostgreSQL with AI-assisted product development. Built and deployed end-to-end analytics applications spanning API ingestion, data quality, KPI investigation and grounded AI insights.

Brings operational leadership and hands-on exposure to sales and service KPIs. Seeking a Werkstudent or internship role across data, business, AI-enabled or product analytics.

VANSH · DATA
STUDENT ID
Portrait of Vansh

Vansh

Data & AI analytics

Studying
B.Sc. DBDS, UE Hamburg
City
Hamburg
Valid till
2028

WHAT I AM

  • Student of Digital Business & Data Science, UE Hamburg
  • Built: AI Workforce Intelligence Platform
  • Building: AI Root Cause Investigation Agent
  • Analysed 41,188 bank-campaign records in Power BI
  • Shift Manager at Domino's, part-time

Vansh

If found, say hello · vanshdhiman090@gmail.com

Quick facts

Location
Hamburg, Germany
Studying
B.Sc. Digital Business & Data Science, 4th semester
Working
Shift Manager, Domino's Hamburg
Languages
English C1 · German B1 · Hindi · Punjabi

AI that ranks and explains. The numbers come from deterministic code.

02 — Skills

What I work with.

01

Data & BI

  • Python
  • pandas
  • NumPy
  • scikit-learn
  • SQL
  • Power BI
  • Excel
  • Exploratory analysis
  • Statistics
  • Forecasting

02

Data Systems

  • PostgreSQL
  • MySQL
  • SQLAlchemy
  • REST APIs
  • ETL / pipelines
  • Data cleaning & validation

03

AI & Development

  • Claude Code
  • ChatGPT
  • Gemini API
  • NVIDIA NIM API
  • Prompt & context design
  • Structured outputs
  • Agentic workflows
  • FastAPI
  • Streamlit
  • Next.js
  • pytest
  • Git / GitHub

04

Soft skills

  • Team coordination

    Coordinate 3-4 team members per shift across baking, service and delivery

  • Working under pressure

    Maintain service standards during high-pressure periods at Domino's

  • Customer communication

    Handle customer escalations and resolve operational issues

  • Attention to detail

    End-of-day cash reconciliation and sales reporting

  • Data-informed decisions

    Review sales, productivity, waiting-time and delivery-time metrics to support shift decisions

  • Multilingual communication

    English C1, German B1, Hindi and Punjabi

  • Self-driven learning

    Built and deployed analytics applications; completed Google and IBM certificates

03 — Selected work

Things I've built.

01AI Workforce Intelligence Platform+

01 · Completed · Live

AI Workforce Intelligence Platform

An end-to-end pipeline that ingests German job postings from the Adzuna API, cleans and deduplicates records, extracts skills and loads normalized data to Neon PostgreSQL.

  • —Deployed dashboard with a Power BI analytics layer
  • —Grounded Gemini summaries
  • —Idempotent refresh
  • —Explicit data-quality disclosures
  • —pytest checks
pythonpostgresqlstreamlitgoogle
02AI Root Cause Investigation Agent+

02 · Active development · Live

AI Root Cause Investigation Agent

A governed workspace for investigating KPI movements using deterministic contribution analysis, bounded hypothesis testing, verification and explicit abstention when evidence is insufficient.

  • —LLM prioritization separated from numerical calculation
  • —323 backend tests
  • —24 browser tests
  • —5/5 controlled benchmark scenarios
fastapinextjspostgresqlgoogleplaywright
03Bank Campaign Intelligence Dashboard+

03 · Completed

Bank Campaign Intelligence Dashboard

Analysis of 41,188 Portuguese bank-campaign records, with a four-page Power BI dashboard covering conversion, customer segments, campaign effectiveness and financial profiles.

  • —11.27% conversion rate identified
  • —High-performing retired and student segments
  • —Near-zero conversion beyond five contact attempts
  • —Findings aimed at targeting decisions
python
04Store-Level Demand Forecasting+

04 · Completed

Store-Level Demand Forecasting

Weekly store-level sales forecasting that compares five methods with walk-forward validation, built for retail and supply-chain planning.

  • —5 forecasting methods compared
  • —Walk-forward validation
  • —Prophet cut MAPE by more than 27% vs a naive baseline
pythonpandas

04 — Certifications

Always learning.

2 certificates

05 — Experience

Education & experience.

Oct 2024 – Feb 2028 (expected)

B.Sc. Digital Business & Data Science

University of Europe for Applied Sciences, Hamburg

4th semester. SQL & relational databases, Python data analysis, applied mathematics & statistics, big data, finance and digital transformation.

Sep 2025

IBM · Python for Data Analysis

Certification

Nov 2025 – Present

Shift Manager / Store Operations All-Rounder

Domino's Pizza, Hamburg · part-time

Coordinates 3-4 team members per shift. Reviews sales, productivity, waiting-time and delivery-time metrics. Handles end-of-day cash reconciliation, sales reporting and escalations.

Mar 2026

Google Data Analytics Professional Certificate

Certification

Next

Your team?

06 — Contact