You will get Sales Forecasting Solution & UI
Project details
You will get a professional demand forecasting solution built with Holt-Winters Exponential Smoothing and Streamlit — ready to use from day one.
I specialize in time series analysis and data science applied to real supply chain challenges. This project was built for a packaging company and delivers automated, intelligent demand forecasts with multi-level aggregation (Corporation, Brand, SKU, Size, Material) and automatic model selection using the AIC criterion.
What you'll receive:
• Accurate forecasts with 95% confidence intervals
• Advanced metrics: RMSE, MAE, and MAPE
• Validation mode to compare model vs. manual PCP forecasts
• Professional Excel export with forecast data + metadata
• Clean, modular Python codebase (Streamlit + Docker-ready)
Whether you need a quick forecast for a single SKU or a full interactive app for your planning team, I have the right tier for you.
I specialize in time series analysis and data science applied to real supply chain challenges. This project was built for a packaging company and delivers automated, intelligent demand forecasts with multi-level aggregation (Corporation, Brand, SKU, Size, Material) and automatic model selection using the AIC criterion.
What you'll receive:
• Accurate forecasts with 95% confidence intervals
• Advanced metrics: RMSE, MAE, and MAPE
• Validation mode to compare model vs. manual PCP forecasts
• Professional Excel export with forecast data + metadata
• Clean, modular Python codebase (Streamlit + Docker-ready)
Whether you need a quick forecast for a single SKU or a full interactive app for your planning team, I have the right tier for you.
Machine Learning Tools
NumPy, pandas, Python, Python Scikit-Learn, scikit-learn, SQL, XGBoostWhat's included
| Service Tiers |
Starter
$300
|
Standard
$450
|
Advanced
$600
|
|---|---|---|---|
| Delivery Time | 3 days | 7 days | 14 days |
Number of Revisions | 1 | 2 | Unlimited |
Number of Model Variations | 1 | 3 | 5 |
Number of Scenarios | 1 | 3 | 5 |
Number of Graphs/Charts | 1 | 3 | 5 |
Model Validation/Testing | |||
Model Documentation | - | ||
Data Source Connectivity | - | ||
Source Code | - | - |
Frequently asked questions
About Giovane
Senior Data Scientist | ML & Causal Inference | MMM | Python | Nubank
Hortolandia, Brazil - 4:56 am local time
🚀 What I Do:
I specialize in end-to-end data science solutions focused on measurable business impact. From causal experiments and marketing analytics to machine learning models and AI agents, I build solutions that are production-ready and grounded in statistics.
📊 Results I've Delivered:
• 2nd place in international AI competition (Bosch, 80+ teams, 1,300+ submissions)
• Winner of Automotive Business Award 2024 — PrognostiQ project (30% improvement in model sensitivity)
• Built MMM (Marketing Mix Modeling), Geo-Tests, and Causal Impact frameworks at Nubank
• Developed propensity models for remarketing campaigns, improving audience segmentation and ROI
• Generated ~BRL 4M in cost savings through ML projects at Messer Gases
• Trained 50+ professionals in Python, SQL, and Power BI
• Built Text2SQL chatbot using RAG + AI Agents for procurement data analysis
🛠️ Core Expertise:
• Causal Inference: MMM, Geo-Tests, Interrupted Time Series, A/B Testing, Incrementality
• Machine Learning: Supervised & Unsupervised, Time Series, Anomaly Detection, Propensity Models
• AI & LLMs: RAG, AI Agents, Text2SQL, LLM integration
• Data Engineering: Databricks, Scala, SQL, Python, Azure ML, MLOps
• Visualization: Power BI, Tableau, Microsoft Fabric
🎓 Education:
Specialization in Data Mining — UNICAMP (one of Brazil's top universities)
Master's-level coursework in Machine Learning & Pattern Recognition — UNICAMP (Grade: A)
🌟 Why Work With Me:
I don't just build models — I align solutions with business goals, communicate clearly with stakeholders, and deliver production-ready code within deadlines. My background in Chemical Engineering gives me a strong analytical mindset and a unique perspective on problem-solving.
Let's build something impactful together. Send me a message and let's discuss your project!
Steps for completing your project
After purchasing the project, send requirements so Giovane can start the project.
Delivery time starts when Giovane receives requirements from you.
Giovane works on your project following the steps below.
Revisions may occur after the delivery date.
Data Validation & Analysis
I review your uploaded Excel file, validate required columns (Company, Brand, SKU, Data, Volume), check data quality, and confirm minimum 12 months of history are present before proceeding.
Model Configuration & Forecasting
I apply Holt-Winters Exponential Smoothing with automatic AIC-based model selection (no trend, additive, or damped trend). Aggregation levels and forecast horizon are configured per your requirements.