The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Back to feedPost
Built a revenue forecasting solution in Snowflake to turn historical operational data into a repeatable forward-looking forecast. I prepared and transformed the data in SQL, engineered time-based and business features, tested multiple forecasting approaches, and used walk-forward backtesting plus forecast-vs-actual analysis to evaluate accuracy, bias, and performance over time.
The network for creativity
Join 1.25M professional creatives like you
Connect with clients, get discovered, and run your business 100% commission-free
Creatives on Contra have earned over $150M and we are just getting started
Related posts
AI-powered financial analytics platform built with FastAPI, Pandas, DuckDB, and Gemini API for data analysis, REST APIs, natural-language-to-SQL queries, and automated business insights.
Most businesses don’t need more AI. They need fewer manual problems.
I work at the intersection of data, AI, and software, turning repetitive workflows, disconnected data, and operational challenges into practical business solutions.
From data analytics and intelligent automation to AI-powered systems, my focus is simple: use technology where it creates measurable value.
Here’s a quick look at what we’re building at D-SAi.
What’s one repetitive process in your business you’d automate if you could?
Title: From Sales Metrics to Cash Flow: Financial Model for a Volatile-Revenue Business
Description:
Built an integrated cash flow and P&L model driven by commercial metrics — sales volume, average ticket size, and customer base growth — rather than pre-aggregated financial figures, for a business with uneven or seasonal revenue.
Challenge: The company could see sales and customer numbers, but had no clear view of how much cash that would actually generate, or when — leaving liquidity planning largely reactive.
Approach: Commercial drivers flow through three linked layers — Commercial Metrics → P&L → Cash Flow — entirely on live formulas, so the model recalculates end-to-end when any input changes. Cash conversion is modeled with a collection-lag assumption (a share of revenue collected the same month, the rest the following month), which is what actually creates the liquidity gap between "profitable" and "cash-positive."
What it surfaced (illustrative figures): In the sample run, EBITDA turned positive by month 3, but the cash balance still dipped to a low point before recovering two months later — the gap between accounting profit and available cash that the business needed to plan around. Every driver (orders per customer, ticket size, collection speed) is a single editable input, so the same model reruns instantly for a different scenario.
Outcome: Turned a sales question into a cash planning answer, with the exact month and depth of the liquidity gap visible in advance instead of discovered after the fact.
