Astha Maurya

Data Engineer

Phagwara, Punjab, India0 mo experience

AI EnabledAI ML Practitioner

Key Highlights

Achieved 96% accuracy improvement in forecasting models.
Developed interactive dashboards for real-time decision-making.
Specialized in end-to-end data pipeline management.

Stackforce AI infers this person is a Data Analyst with a strong focus on Machine Learning and Data Visualization in the Energy sector.

Contact

Skills

Core Skills

Machine LearningData AnalysisData Visualization

Other Skills

TF-IDFNLPFlaskCosine SimilarityDjangoUser AuthenticationDatabase AnalyticsPower BIDAXData CleaningData TransformationArtificial IntelligenceProject-Based LearningScikit-LearnProphet

About

I don't just analyze data — I build systems that make data useful. As a B.Tech CSE student at Lovely Professional University (2027), I specialize in the complete data pipeline: from wrangling multi-state electricity datasets and engineering time-series forecasting models, to deploying production-grade interactive dashboards that real stakeholders use for decision-making. Here's what I've shipped: 96% accuracy improvement — built and benchmarked 3 forecasting models (Prophet, Linear Regression, Naïve) on India's energy data (1985–2023); Prophet achieved MAPE of 4% on Coal vs Naïve RMSE of 87,358 · Edunet Foundation × AICTE × Shell 476-row electricity demand dataset across 15+ Indian states processed with 100% data integrity; AI-driven forecasting model predicting 3–12 month supply trends via Facebook Prophet · REC Limited Movie Recommendation Engine — NLP-powered content-based system using TF-IDF + cosine similarity on 5,000+ TMDB movies, served via Flask REST API 15+ visualization Power BI dashboard on India Air Pollution — 1,199 records, DAX-driven risk classification across High/Moderate/Low zones My stack: Python · SQL · Power BI · Streamlit · Scikit-learn · Prophet · Flask · Django · Groq AI API · NLP. I'm actively seeking Data Analyst / Data Scientist / ML Engineer internships where I can build real things, not just shadow senior teams. If your org runs on data and needs someone who can both analyze it and deploy it — let's connect.

Experience

0 mo

Total Experience

Average Tenure

Current Experience

Github

3 roles

System Specialist

Feb 2026 – Mar 2026 · 1 mo

Built a content-based recommendation engine using TF-IDF vectorization and cosine similarity on a 5,000+ movie TMDB dataset, incorporating weighted metadata (genres, cast, director). Applied NLP preprocessing — tokenization, Porter Stemming, stop-word removal — to build enriched tag vectors, served via a Flask REST API returning top-5 recommendations. Pre-computed similarity matrices and persisted model artifacts using Pickle, enabling instant inference and significantly reduced response latency.

TF-IDFNLPFlaskCosine SimilarityMachine LearningData Analysis

Learning Specialist

Jan 2026 – Feb 2026 · 1 mo

Created a full-stack AI-powered learning platform using Django with user authentication, quiz sessions, and database-driven analytics to track performance across multiple aptitude topics. Integrated the Groq AI API into an adaptive quiz engine that dynamically generates 5+ questions per session based on user performance. Built a performance analytics module classifying topics into 3 tiers (Strong, Moderate, Weak) for targeted personalized practice.

DjangoUser AuthenticationDatabase AnalyticsMachine LearningData Analysis

Analysis Specialist

Nov 2025 – Dec 2025 · 1 mo

Evaluated 1,199 air quality records from monitoring stations across Indian states, performing data cleaning, transformation, and pollutant normalization using Power Query. Designed a multi-page interactive Power BI dashboard with 15+ visualizations and KPI indicators, enabling analysis of pollution hotspots, pollutant distribution, and state-wise severity trends. Integrated DAX-based measures to categorize cities into High, Moderate, and Low pollution risk zones, supporting public health awareness and environmental analysis.

Power BIDAXData CleaningData TransformationData VisualizationData Analysis

Edunet foundation

2 roles

Intern

Jan 2026 – Feb 2026 · 1 mo · Remote

Artificial Intelligence & Machine Learning Intern
Edunet Foundation | AICTE | IBM SkillsBuild
📅 Jan 2026 – Feb 2026 (Ongoing)
Selected for a 6-week AICTE–Edunet Foundation internship under the IBM SkillsBuild Program, focused on Artificial Intelligence and Machine Learning.
Currently learning and applying AI & ML fundamentals through structured e-learning modules and mentor-led technical sessions.
Working independently on a project-based problem, guided by an assigned mentor, to develop a real-world solution using AI/ML concepts.
Actively participating in weekly technical sessions, Q&A discussions, and masterclasses conducted by subject matter experts.
Enhancing skills in problem-solving, self-paced learning, project documentation, and presentation, with a final project submission planned.
Skills: Artificial Intelligence · Machine Learning · IBM SkillsBuild · Project-Based Learning · Data Analysis · Problem Solving · Technical Documentation

Artificial IntelligenceMachine LearningProject-Based LearningData Analysis

Intern

Oct 2025 – Nov 2025 · 1 mo · Remote

Analyzed India's electricity generation data across 38 years (1985–2023) and 8 energy sources (TWh), achieving 100% data completeness with zero missing values
Built and benchmarked 3 forecasting models (Prophet, Linear Regression, Naïve) — achieved best MAPE of 4% on Coal; Prophet RMSE of 5,518 vs Naïve RMSE of 87,358, a 96% accuracy improvement
Delivered an interactive Streamlit dashboard with 4+ KPIs and Plotly visualizations tracking CO₂ emissions, renewable energy growth, and energy mix trends for sustainability planning

Data AnalysisScikit-LearnProphetLinear RegressionMachine Learning

Edunet foundation (aicte + shell)

Data Specialist

Oct 2025 – Nov 2025 · 1 mo

Analyzed India's electricity generation data (1985–2023) across 8 energy sources (TWh), achieving 100% data completeness with zero missing values for time-series forecasting. Developed and compared 3 forecasting models (Prophet, Linear Regression, Naïve) — best MAPE of 4% on Coal; Prophet RMSE of 5,518 vs Naïve RMSE of 87,358 — a 96% accuracy improvement. Built an interactive Streamlit dashboard with 4+ KPIs and Plotly visualizations to track CO₂ emission trajectories, renewable energy growth, and energy mix trends for sustainability planning.

Scikit-LearnProphetData AnalysisMachine Learning

Rec limited

Data Analyst

Jul 2025 – Aug 2025 · 1 mo · Gurugram · On-site

Processed 476 rows × 6 columns of multi-state electricity demand data spanning 15+ Indian states with 100% data integrity, performing EDA to engineer a forecast-ready dataset for power consumption analysis
Engineered an AI-driven time-series forecasting model using Facebook Prophet (mean: 4,083 units) to predict 3–12 month electricity supply trends, enabling state-level comparison of energy requirement vs energy delivered
Deployed a production-grade Streamlit dashboard with authentication, forecast history tracking, dynamic filters, forecast confidence intervals, and KPI summaries — actively used for energy planning decisions

ProphetScikit-LearnData AnalysisMachine Learning