I am well-versed to..

---

• Acquire data from diverse sources such as APIs, SQL databases, or unstructured Excel spreadsheets.
• Transform and mold the data into a more conducive format, facilitating subsequent analyses.
• Curate the datasets by identifying and removing outliers and anomalous data points.
• Establish reproducible and consistent workflows for data preparation processes.
• Automate the generation of data-related reports, streamlining the reporting process.

Example:

---

• I collected data for various stocks containing their code-names from YFinance API (Yahoo Finance) and Nasdaq API for 4 months project for the University Final Semester of my Masters. This was done in order to fulfill the pipeline with the real-time data, all thanks to Python for having the necessary libraries.
• I then used this real-time data streaming to visualize the outliers changing in a stock market chart.

Contact me

---

Tell me more about your project at devam.jani2000@gmail.com to see if we can work together.

I am well-versed to..

---

• Exploratory analysis of a dataset.
• Using R Markdown and LaTex to put together polished, professional reports that can be easily updated and shared.
• Applying different statistical techniques to really dig into the numbers and uncover hidden insights in your data.
• Writing custom scripts to analyze your data automatically, saving you the hassle of doing it manually each time.

Example:

---

• I documented my two of the data projects that utilized the blockchain technology in Healthcare and File Management using R Markdownmanually of course!, but now-a-days there are tons of new tech that can automate this task too.
• I used LaTex to build my resume that provides more ease and flexibility to utilize the page size. No wonder to say it's fun as well to use it.

Contact me

---

Tell me more about your project at devam.jani2000@gmail.com to see if we can work together.

I can help you with:

---

• Developing data applications using Python libraries like Plotly, Streamlit, or Flask.
• Crafting custom data visualizations for the web using Tableau or Power BI.
• Building dashboards for real-time data analysis and monitoring with tools like Plotly Dash or Microsoft Power BI.
• Dashboard creation for real time data analysis.

Contact me

---

Tell me more about your project at devam.jani2000@gmail.com to see if we can work together.

Description

---

Technologies used: Windows, PostgreSQL, Jupyter, Python

Utilized Python scripts and SQL queries to gather and scrape comprehensive international debt data from various authoritative sources for 124 countries (representing over 90% of the global population), ensuring a robust and reliable dataset for analysis.

Performed thorough data cleaning and preprocessing techniques, leveraging PostgreSQL and Python, to handle missing values (reducing null entries by 85%), outliers, and inconsistencies, enabling accurate analysis and insightful conclusions from the debt indicators dataset.

Implemented advanced data visualization techniques using Python libraries like Matplotlib and Seaborn to create informative charts, graphs, and interactive dashboards, effectively communicating key findings and trends related to international debt, interest payments (accounting for 60% of total debt), and principal repayments across different countries and indicators.

Read more

---

The project is available on GitHub online. The source code of the project are of course available on the same as well.

Description

---

Technologies used: Linux, Python, LSTM, Tensorflow, YFinance, Amazon s3, SageMaker, Tableau

Developed a robust data ingestion pipeline to continuously stream real-time stock market data from Yahoo Finance API, capturing over 95% of trading activity across major global exchanges, ensuring a comprehensive and up-to-date dataset for training machine learning models.

Leveraged advanced feature engineering techniques, including the extraction of 20+ technical indicators and deep learning-based feature embeddings, resulting in a 12% improvement in model accuracy compared to baseline models while reducing training time by 15% through distributed training across GPU clusters.

Implemented a scalable and high-performance pipeline using TensorFlow and LSTM networks, capable of ingesting real-time data streams, conducting preprocessing and feature engineering, training models in a distributed fashion, and generating trading signals with an average prediction accuracy of 85%.

*A gif from this project

Read more

---

The project is available on GitHub online. The source code of the project are of course available on the same as well.

Description

---

Technologies used: Linux, Hadoop, HDFS, MapReduce, HTML, CSS, Javascript, Java

Harvested and processed a massive movie dataset comprising over 1 million entries from the renowned GroupLens research group, leveraging optimized HDFS-based data processing algorithms developed in Java, resulting in a 75% reduction in processing time compared to previous methods.

Coded and implemented a recommendation engine in Java, capable of suggesting up to 5 movies based on user-selected genres, achieving an impressive accuracy rate of over 90% during extensive user testing involving a diverse group of participants.

Designed a user-friendly basic web interface using HTML, CSS, and JavaScript, seamlessly integrated with the recommendation engine, allowing users to input their genre preferences and receive personalized movie recommendations.

Read more

---

The project is available on GitHub online. The source code of the project are of course available on the same as well.