You will get Data Analysis, Dashboarding & Statistical/Econometric Insights
Project details
I turn raw, messy data into clear analysis and visuals that support real business decisions. With training from WorldQuant University's Applied Data Science Lab and IBM's Machine Learning certification, I combine statistical rigor (regression, time series, panel data) with clean, interactive dashboards, so you get both the "what's happening" and the "why it matters."
Data Tool
Microsoft Power BIWhat's included
| Service Tiers |
Starter
$50
|
Standard
$100
|
Advanced
$200
|
|---|---|---|---|
| Delivery Time | 3 days | 6 days | 8 days |
Number of Revisions | 1 | 2 | 3 |
Number of Pages Mined/Scraped | 0 | 2 | 5 |
Number of Sources Mined/Scraped | 1 | 2 | 3 |
Optional add-ons
You can add these on the next page.
Fast Delivery
+$50 - $150
Additional Revision
+$50
Additional Page Mined/Scraped
(+ 4 Days)
+$100
Additional Source Mined/Scraped
(+ 3 Days)
+$150Frequently asked questions
2 reviews
(2)
(0)
(0)
(0)
(0)
This project doesn't have any reviews.
DS
Don S.
Aug 3, 2026
Data Automation Expert for Bank Reconciliation
Built a highly efficient, automated tool with smart AI handling for matching edge cases. Zero errors, unbelievable speed, and immaculate setup. Easily one of the top freelancers on the platform. Looking forward to our next project together!
UU
Usonwanne U.
Aug 3, 2026
Python Data Analyst for Financial Time-Series Modeling
Outstanding quantitative analyst! Handled panel data cleaning, outlier treatment, and advanced time-series modeling (VAR) with absolute precision. Their econometric expertise, clean Python code, and crystal clear presentation of forecast scenarios exceeded all expectations. If you get the chance to hire him, do it immediately.
About Perryman
Expert Machine Learning Engineer| Forecasting | Data Science| Python
100%
Job Success
Fresno, United States - 2:27 pm local time
I am a machine learning engineer and an Economist who works across the full analytical pipeline, from raw messy data to deployed production system. My two most recent projects are not portfolio screenshots, they are live systems anyone can open in a browser right now.
MacroSense is a production Machine Learning forecasting system that predicts US GDP growth, inflation, and unemployment six months ahead using Federal Reserve data. Built across a rigorous six-stage pipeline covering API data collection, exploratory analysis, feature engineering with stationarity transformation and lag variables, ensemble modelling, and walk-forward validation across 25 years of economic history. Directional accuracy of 88.9% for GDP forecasting. Live on Streamlit Cloud.
CreditIQ is a loan default prediction system trained on 150,000 real borrower records. It goes beyond standard model comparison by implementing business cost threshold optimisation, finding the exact decision boundary that minimises total dollar loss for the lender rather than just maximising a statistical metric. Ensemble AUC-ROC of 0.844 with full SHAP explainability for every individual prediction. Also live on Streamlit Cloud.
My third published project, a comparative study of OLS econometrics versus machine learning for Nigerian GDP forecasting, produced an 8x improvement in predictive accuracy using XGBoost over traditional regression and was submitted to the Journal of Economics and Business Review in 2025.
What I bring is the combination of formal economics training and production Machine Learning engineering. An economics background means I understand the domain behind the data, why a variable matters, whether a finding makes economic sense, and when a statistically significant result is economically meaningless. That understanding shapes every modelling decision I make before writing a single line of code.
I can support your project across the following:
-- Data extraction, cleaning, matching, validation, and pipeline preparation
-- Machine learning, classification, regression, clustering, anomaly detection, and predictive analytics
-- Machine Learning Time series forecasting using ARIMA, GARCH, XGBoost, and ensemble methods
--Machine Learning Financial analysis: credit risk modelling, and business cost optimisation
-- Econometrics: panel data, causal inference, difference-in-differences, propensity score matching, and instrumental variables
-- Statistical analysis and hypothesis testing using Python, R, and Stata
-- Excel and Power BI dashboards and automated reporting for non-technical stakeholders
-- NLP, sentiment analysis, text classification, and qualitative content analysis
-- SHAP explainability and interpretable AI for regulated or presentation-facing outputs
-- Machine Learning End-to-end deployment on Streamlit Cloud with live production systems
Across all of this the standard is the same: clean documented work, honest evaluation, and outputs that hold up when someone asks hard questions about how the numbers were produced.
My deliverables typically include organised datasets, clean and reproducible code, documented assumptions, validation checks, dashboards or reports, and a concise summary of the main findings.
I communicate regularly, identify data limitations early, and focus on solutions that are accurate, practical, and easy to maintain.
Send me your data or project requirements, and I’ll suggest the most effective approach.
Steps for completing your project
After purchasing the project, send requirements so Perryman can start the project.
Delivery time starts when Perryman receives requirements from you.
Perryman works on your project following the steps below.
Revisions may occur after the delivery date.
Kickoff & data review
— I review your dataset and confirm the business question you want answered
Data cleaning
— Handle missing values, duplicates, and inconsistencies to ensure accuracy