I will build machine learning models for forecasting or prediction
Data Analyst I SQL, Python, Power BI, Tableau I Predictive Analytics, Dashboards
About this Gig
I build machine learning models that turn your historical data into forward-looking predictions sales forecasts, customer churn risk, demand planning, or classification problems specific to your business.
As an AI Engineer with professional experience deploying production ML systems, I bring real-world modeling experience beyond textbook examples. I focus on models that are accurate, explainable, and actually usable by your team not just impressive-looking accuracy scores.
Whether you need a one-off prediction model or a documented, reusable pipeline, I'll match the solution to your data and business goal.
Why choose me (ML-specific don't reuse from other gigs):
- Professional experience building and deploying real ML systems, not just tutorials
- MSc in Big Data Analytics with strong grounding in statistical modeling
- Models delivered with clear documentation you'll understand what it does and why
- Experience across classification, regression, clustering, and time-series forecasting
You can see my previous projects and dashboards here:
GitHub link: https://github.com/Chooladeva
Portfolio website link: https://sites.google.com/view/chooladeva-piyasiri/portfolio?authuser=0
My Portfolio
FAQ
What kind of predictions can you build models for?
Sales/demand forecasting, customer churn, risk or credit scoring, and general classification or regression problems. If your use case is different, message me to check fit before ordering.
What format should my data be in?
CSV, Excel, or a SQL export. Upload a sample when placing your order so I can confirm what's feasible.
Will I understand how the model works?
You'll get a plain-language summary of what factors are driving the predictions — not just a script with no explanation.
Can you deploy the model somewhere I can use it directly?
Yes, on the Premium package — I can wrap it in a simple Streamlit app so you can run new predictions without touching code.
Do you handle missing or messy data?
Yes — data cleaning and preprocessing is included as part of the modeling process in every tier.
How accurate will the model be?
Accuracy depends heavily on your data quality and size — I'll give you an honest assessment upfront rather than promise a specific number before seeing your data

