My work involves deriving insights and business use cases leveraging multidimensional datasets, developing data processing pipelines, prototypes and ML models, researching opportunities and producing realistic project plans, and delivering satellite data-based and diverse data-centric projects to clients- spearheading the development of satellite data-based projects as well as contributing to other ML projects, utilizing high-quality data and ensuring the delivery of accurate, reliable, and actionable insights to drive innovative development- Built data acquisition pipelines using both the Sentinel-Hub API and Google Earth Engine, streamlining data collection for various geospatial tasks and map dashboards- Carried of research and development of a key-company product - Bark beetle detection, forest vegetation monitoring and digital MRV solutions- Developed a deep learning model for tree cover detection. Leveraged tree data and allometric equations to create an end-to-end predictive model for estimating carbon stocks. Utilized biophysical variables extracted from satellite data, improving the accuracy of carbon sequestration assessments for client-defined field boundaries.-Developed in-house processing pipelines and analysis tools for working with Sentinel-2, Planet and SAR data (Sentinel-1) for extracting biophysical indices for various core products- Database design, Data Extraction / Transformation / Load (ETL) in Open Telekom Cloud- Rapid prototyping and implementation of data models and DL models- Advanced image processing and Deep learning applied to diverse image formats- Scientific coordinator and project manager for 4 interns in Data science and engineering- Documentation, reviewing and implementing methods from research papers, writing business proposals, and report summarizing methodologies and results- Supported various web app developments and company tasks by deploying APIs, managing databases, and creating dashboards

- Developed ETL pipeline and automated the processing of large-scale AIS (Automatic identification systems) to store them OTC cloud - PostgreSQL- Analyses EO data from all Copernicus satellites to develop new business cases

- Created ETL processes to clean, analyze, visualize, and extract insights from call detail record datasets from multiple cellphone towers in the US-Employed K-means clustering and evaluated Autoregressive Integrated Moving Average (ARIMA) model on time series data to predict anomalous behavior of the network and subscribers with high accuracy

- Outlined work plans, allocated tasks, timelines and updated weekly progress- Coordinated with other project managers, organized regular meetings and feedback sessions- Prepared project progress reports and communicated overall weekly progress via engaging presentations to professors, tutors and other members of the project

- Developed system architecture for components of ISS look-alike- Data analysis of sensor data from spacecraft components to optimise - launch, manoeuvre and landing of spacecrafts- Assessed crew capabilities and psychological aspects of interplanetary travel

My master thesis was focused on processing and analysis of multi-dimensional astronomical data to look for the presence of astrophysical signals and to study the nature of Pulsars (rotating neutron stars) and Fast radio bursts through single pulse search techniques. Tasks and responsibilities:- Developed a data processing python pipeline that processed 9 TB of raw radio data (Astrophysical observations) on a Linux based supercomputing cluster to look for astrophysical signals : Data reduction, de-noising, Fourier transform, box-car smoothing, data visualization.- Data processing was complimented by bash wrappers to python scripts backed by an SQL database that facilitated analysis and automated data reduction- Wrote SQL queries to extract information of different features (frequency, source name, etc.) of each data file and analyzed data to find correlations between features- Developed an interactive application using python to visualize a 3D image (Dynamic Spectra) of the pulse from an astrophysical source.- Tested and validated Machine learning models (linear regression, Clustering, CNN, SVM) for the classification of Astrophysical signals-Statistical modelling of Pulsar energy distribution- Presented reports in group meetings and effectively communicated with the project supervisor and came up with creative solutions to handle data and derive conclusions.Skills acquired: Python, Bash, SQL, numpy, pandas, matplotlib, scikit-learn, statsmodels, OpenCV, Excel, report writing, presentations, Docker.

Tasks and responsibilities: -Developed an algorithm based on the idea of Hough transform (feature extraction technique) to detect pulsars (rotating neutron stars) and prepared data to train an Neural network- Performed statistical analysis on the data to understand the correlations between different featuresSkills acquired: Python, Bash, statistical analysis, training neural network, Docker.

Tasks:Worked on manual classification and visualization of the properties of known neutron stars to understand population statistics and evolutionSkills Acquired: Python, statistics.