← TEAM / DISCIPLINE 01PYTHON →
JOB SUCCESS100%Upwork track record STATUSTop Rated PlusUpwork recognition CLIENT HOURS1,375+Documented hours UPWORK JOBS4Total jobs Python Web AppsPython · SQL · HTML/CSS/JSOperate BI DashboardsPower BI · Looker Studio · ExcelDecide Data PipelinesCSV · Excel · XML · DatabasesAutomate
Welcome to Morage's workspace.
PYTHON →
PRODUCT.
Data Science Professional and Python Developer building clear dashboards, useful web tools, automation and data-connected systems.
UPWORK SNAPSHOT · SEP 2026
Professional activityUPWORK / CLIENT WORK
PROFESSIONAL SIGNAL
Top Rated Plus freelancer with 100% Job Success and 1,375+ documented client hours, working across Python development, data science and business-facing systems.
BuildStackUse
PROFILE / 01CODE DATA.
CODE DATA.
SHIP TOOLS.
Morage is a Data Science Professional and Python Developer focused on turning complex information into clear analysis and practical digital tools. Her work combines statistical analysis, predictive modelling, dashboard design and full-stack Python workflows.
She builds backend systems, admin panels, file-processing automations and database-connected web applications while keeping the user experience clear enough for real teams to operate.
TECHNICAL STACKPYTHON,
PYTHON,
CONNECTED.
PythonJavaScriptHTML5CSS3Admin PanelsForm Automation
PandasScikit-learnStatistical ModelingRMSE / MAERSQL
Power BILooker StudioExcelKPI DesignExecutive Reporting
MySQLMongoDBCSV / Excel / XMLGitGitHubBackend Integration
SELECTED CLIENT WORKCLIENT WORK.
CLIENT WORK.
DELIVERED.
Python Projects
Signovus Capital LLC
ONGOING DEVELOPMENT PARTNERSHIPLong-term Python project work spanning practical automation, data handling and software workflows. Presented as an ongoing engagement rather than a one-off portfolio item.
Decoding Pything and Security check of Python program- Very complicate project
Upwork
★★★★★5.0 / 5.0 · CLIENT REVIEW
“Reliable, Honest, Dedicated, Patient! highly recommended”
Python Developer Needed
Upwork
★★★★★5.0 / 5.0 · CLIENT REVIEW
“It was a pleasure working with Morage on the Python project. Morage demonstrated excellent communication skills throughout the project, making collaboration seamless and efficient. He quickly grasped the technical requirements and nuances of the task, showcasing his ability to learn rapidly and adapt to new challenges. His performance exceeded expectations, and I highly recommend him for any future projects.”
SELECTED PORTFOLIO / WORKFLOWPYTHON FIRST.
PYTHON FIRST.
BUSINESS READY.
A connected workflow from messy information to a usable data product. Each stage shows how Python, analysis and business intelligence fit together rather than treating them as separate tools.
RAW DATA → PYTHON → ANALYSIS → DECISION → PRODUCTRaw Data
Bring CSV, Excel, XML and database inputs into a clean structure that can be checked, transformed and reused.
Python Systems
Validate, transform and automate repeated work with Python logic, data processing and API-connected workflows.
Data Science
Explore patterns, compare models and evaluate results with transparent metrics before choosing what actually generalises.
Business Intelligence
Turn analysis into executive-ready dashboards, KPI views and reporting that make the next decision easier to see.
Interactive Products
Package the workflow into practical web tools people can operate — including EVA DataLab for guided small-data analysis and decision support.
Launch EVA DataLab ↗DATA SCIENCE / CASE STUDIESMODEL THE PROBLEM.
MODEL THE PROBLEM.
THEN TEST THE MODEL.
Two end-to-end studies showing Morage’s approach to exploration, feature engineering, model comparison and honest evaluation on unseen data.
MovieLens /
Large-Scale Recommendation Modeling
Explored a large ratings dataset, engineered user-, movie-, time- and genre-level features, then compared baseline, regularized, SVD and Spark ALS approaches. The final selection prioritised holdout performance, scalability and generalisation rather than the best-looking validation score.
RATINGS9.0M
USERS69,878
MOVIES10,677
ALS HOLDOUT RMSE0.8177
SIMPLE AVG≈0.91
REGULARIZED≈0.88
SVD / HOLDOUT3.6661
SPARK ALS0.8177
KEY DECISION — SVD looked strong on validation but deteriorated sharply on the final holdout set. Spark ALS became the final model because it combined the lowest holdout RMSE with better sparse-data scalability.
Open case study PDF ↗Liver Disease Prediction /
Sensitivity vs Specificity
Prepared the Indian Liver Patient Dataset through missing-value treatment, encoding, feature-importance analysis and model comparison. Rather than selecting the highest headline accuracy, the study examined the trade-off between finding positive cases and avoiding false positives.
RECORDS583
VARIABLES11
K-NN SENSITIVITY81.45%
K-NN SPECIFICITY36%
Decision Tree100% sens.0% spec.
Random Forest89.52% sens.26% spec.
K-Nearest Neighbors81.45% sens.36% spec.
Logistic Regression99.17% sens.1.85% spec.
KEY DECISION — K-NN was selected in the report for the more balanced sensitivity/specificity profile. The study also flags class imbalance, scaling, outlier sensitivity and validation as important limitations.
ACADEMIC / PORTFOLIO MACHINE-LEARNING STUDY — NOT A CLINICAL DIAGNOSTIC TOOL.Open case study PDF ↗CERTIFICATIONS / LEARNINGPROOF,
PROOF,
NOT BADGES.
Professional Profiles
See Morage’s live professional profiles for additional certifications, credentials, current work history and projects.
LANGUAGESCOMMUNICATE / BUILD / DELIVER
EnglishNative or Bilingual
CantoneseNative or Bilingual
ChineseNative or Bilingual
UrduNative or Bilingual
Panjabi / PunjabiFluent
PROJECT INTAKE / 24-7Need a Python developer who understands the data too?Start with the workflow, the dataset or the business problem.
Start a project ↗