Priyansh Singh Nirwan

Results-driven Data Analyst with experience at Samsung Research and Development, skilled in Python and advanced SQL. Successfully built ETL pipelines and designed data visualization dashboards, enhancing IoT device performance by 20%. Adept at bridging technical gaps and collaborating with stakeholders to implement impactful solutions.

• 26th April, 1999

• Best Collaboration Award : Collaborated with the SRCN (China) team to analyze IoT device data, awarded by Samsung Research Institute, Delhi.
• Best Innovation Award : Developed a system performance evaluation framework using Python and Matplotlib, awarded by Samsung Research Institute, Delhi.
• Team Leadership : Led a team of 3 members to deliver a defect prediction model during the integration phase, ensuring collaboration and timely execution.

Yulu - hypothesis testing

1. Analyzed factors affecting electric cycle demand using EDA and statistical tests on more than ten variables

2. Performed t-tests and ANOVA to identify variables impacting rentals 

3. Conducted chi-square tests to analyze weather season dependency

4. Improved rental prediction accuracy by 15% through data-driven insights 

Skills: Python, Pandas, Matplotlib, SciPy, hypothesis testing, data visualization

Location-based recommendation system

Developed a recommendation system integrating location-based services to suggest essential goods tailored to users' locality 

Implemented collaborative filtering, content-based filtering, and hybrid models for personalized recommendations 

Utilized algorithms like cosine similarity, singular value decomposition (SVD), and non-negative matrix factorization (NMF)

Improved recommendation accuracy with SVD and NMF models, achieving RMSEs of 1.19 and 1.29, respectively 

Skills: Python, Scikit-learn, Jupyter Notebook, Recommendation System Design, and Statistical Analysis

Digital Sentinel: Cybercrime analytics

Designed and implemented an ETL workflow to preprocess and aggregate cybercrime data 

Built predictive models using classification algorithms (logistic regression, random forest) to categorize risks with an accuracy of 85% 

Analyzed trends in phishing, fraud, and ransomware attacks to provide actionable insights for preventive measures. Skills: Python, Pandas, Scikit-Learn, Matplotlib, Seaborn, Predictive Analytics