Urgent.News

What's breaking now, across thousands of outlets.

AI

I Built ScopePilot: An AI-Powered Proposal Generator Using Gemma 3 That Runs Locally

Freelancing isn't just about doing the actual work. Before a project even begins, freelancers often spend hours understanding client requirements, defining deliverables, estimating costs, planning milestones, and writing professional proposals. And the problem? A client brief is rarely perfect. Sometimes it's just a few sentences in an email. Sometimes it's a long document with unclear…

Creating proposals is a crucial yet time-consuming task for freelancers. Often, client briefs are brief and unclear, requiring significant effort to turn them into a formal proposal. The goal was to build an AI-powered assistant to simplify this process. Introducing ScopePilot - a local-first AI tool that generates structured proposals from unstructured client briefs.

ScopePilot is designed for freelancers, independent developers, small agencies, and those new to freelancing. When given a client brief, ScopePilot generates: a project analysis, executive summary, scope and deliverables, technology recommendations, milestones and timeline, pricing breakdown, risks and assumptions, and clarification questions.

These details can be edited, regenerated, saved, and exported as PDF or Markdown files. The core AI engine powering ScopePilot is Gemma 3 1B, which runs locally through the Ollama platform. The application communicates with Ollama via its local REST API. Validating the model's outputs, handling finances, and managing local storage are all managed by Python services.

The project is open-source, allowing users to run ScopePilot locally without sending client information to external AI services. This approach prioritizes privacy, cost-effectiveness, and user control. The development process taught valuable lessons about integrating AI models into applications, where careful design of the surrounding infrastructure is just as important as the AI itself.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in AI

Stuck in an AI fix loop? Do this before the next prompt.

You know the feeling. Something breaks. You ask the AI to fix it. It says "Fixed!" It isn't. You paste the error again. It apologizes and says "Fixed!" again.

  • Stop repeatedly asking AI to fix same issue; reset and take a break.
  • Write down four key details: action, expectation, outcome, error message.
  • Start new chat with concise summary if conversation becomes tangled.

More from Sunday 18 October →