Forecasts & customer intelligence

E-commerce and small businesses with a few years of orders and no data team.

Your sales history already knows what will sell next month and which customers are drifting away. Getting that out takes clean data, explainable models and a dashboard people will actually open.

• +Data audit and cleanup of your orders, customers and products
• +Sales forecasting at the horizons you plan around
• +RFM segmentation and churn-risk scoring
• +Models served from an API you can retrain on new data
• +A dashboard — or the numbers wired into the tools you use

• Python
• Prophet
• scikit-learn
• pandas
• FastAPI
• PostgreSQL

I'll work with your existing stack where it makes sense — these are defaults, not requirements.

SellWise · Live · 2026

Forecasts and customer intelligence for small online shops.

1. 01 · Data

   Orders, customers, stockPostgreSQL · CSV import
2. 02 · Code

   Node.js REST APIOrders, inventory, invoices, reports
3. 03 · Model

   Demand forecastProphet · 7 / 14 / 30 days
4. 04 · Model

   Segments & churnRFM · logistic regression
5. 05 · In / out

   DashboardForecasts, customer risk, stock alerts

1. 01

   Intro call

   30 minutes, free. You walk me through the workflow; I ask about your data, your systems and what “done” means.

2. 02

   Written scope

   A short document: what gets built, what's explicitly out, milestones, and a price. No surprises later.

3. 03

   Build in the open

   A demo on a live staging URL every week. You see working software, not status reports.

4. 04

   Handover

   Your repository, deployment, documentation and tests. Support after launch if you want it.

• →Deterministic where it must be right. Money, permissions, routing and anything with consequences is decided by plain code that can be tested and audited. Models read, interpret and draft — they don't get the final say.
• →Guardrails are code, not prompts. A system prompt is a request. Validation, schema checks and post-processing are guarantees. Every model output is checked before a user or a database sees it.
• →It keeps working when the model doesn't. Providers time out, keys expire, outputs drift. I design a safe fallback path for each model call so the product degrades gracefully instead of failing loudly.