02 Experience

Education

Chapter 01 of 03

Jan 2026 – Present · Hamburg, Full-time

Built the complete news-ingestion pipeline: extracts news from GDELT in real time, embeds it, and stores it in AWS S3 vector buckets.
Two pipelines run in parallel — one scans GDELT for news as it breaks, the other processes and stores historical data at the same time.
Built a retrieval library on top: vector search with ranking for the most reliable results, plus filters for dates, countries, and more.
Developed an internal data-production tool for our analysts: one automated pipeline from data processing to forecasting, KPI derivation, and publishing.
Current focus: turning the survey data team's most time-consuming manual workflows into automated Python pipelines — and building them an internal operations app with a proper UI (React front end, Python back end) that runs the whole pipeline.

Apr 2025 – Jan 2026 · Hamburg

Owned the Public Cloud, Semiconductors, AI, and Data Center markets — forecasting and publishing forecasts, KPIs, and reports in Statista's Market Insights.
Introduced new modelling methods with a real-time news indicator as the headline feature: forecasts reflect what is actually happening in the news.
KPI coverage included revenue, market shares, and tariff changes; introduced new segments across these markets.
Helped clients with their data and methodology requests.

Dec 2023 – Mar 2025 · Hamburg

Handled the Data Center and Public Cloud markets.
Produced forecasts and published forecasts, KPIs, and reports in Statista's Market Insights.

Jul 2022 – Oct 2023 · Hamburg

Built web scrapers to collect data and convert the scraped data into structured data.
Explored Common Crawl data — extracted the useful parts, converted them into structured data, and stored them.
Used NLP with Google's BERT model to find similar data across multiple datasets and compare how it's presented differently — e.g. product data across multiple stores.

Mar 2025 – Present · Freelance, Remote

Designed AI-powered extraction workflows for image and document processing use cases.
Built production-oriented tools for OCR, structured JSON extraction, and multi-provider LLM automation.
Delivered applications focused on secure processing, operator usability, and business-ready outputs.
Sole engineer and client contact — scoping, system design, shipping, and ongoing maintenance of a platform used daily by ~200 people.

Jul 2020 – Jan 2021 · Chennai, India

Worked on data preparation, visualization, statistical analysis, and machine learning assignments.
Identified the key influencing features affecting drone battery performance.
Identified and reported drone issues during flight for the engineering team.

Mar 2023 – Oct 2023 · Germany

Developed a classifier using Convolutional Neural Networks (CNNs) to identify common plant and insect species in a specific region of Germany.
Collected and curated a large dataset of plant and insect images through web scraping, followed by data cleaning and preprocessing.
Designed and implemented an API for the classifier — users submit images and receive species identification results.
Built a continuous-training pipeline, ensuring consistent performance improvements and adaptability to new data.

M.Sc. Smart Systems · 2021 – 2023

B.E. Mechatronics Engineering · 2017 – 2021 · Grade 8.27 / 10