CAPSTONE RESEARCH · LIMITED PREVIEW

Find a capstone idea you can actually build.

ScoutSpark helps students explore ideas that fit their interests, skills, time, and resources. It searches for supporting sources and shows when evidence is not strong enough for a detailed plan.

Research runs locally with your own provider setup. This Pages site is a static project overview.

A RESEARCH PATH YOU CAN FOLLOW

From question
to evidence.

Move through a reviewed workflow. See what was searched, how ideas were assessed, and where the evidence is still limited.

01

Scope a project

Set the domain, time available, interest area, and the kind of idea you want to explore.

02

Choose Live sources

Use configured search providers for current sources, or select Offline demo to explore synthetic examples.

03

Review the run plan

Preview the resolved sources, limits, scoring weights, and budget before research starts.

04

Inspect ideas and evidence

Compare scorecards, gate outcomes, and source receipts. A high score does not clear a failed gate.

SEE THE FLOW

Live mode, step by step.

00:30 16:9

Download the MP4

01 Live sources02 Scope the request03 Preview the plan04 Inspect results05 System overview

LIMITED PREVIEW

Evidence first. Review every result.

Results can be incomplete or unverified. Check sources and warnings before relying on an idea. A small set of evaluations has shown real quality and proposal-yield limitations.

Read the preview notes

ONE SYSTEM VIEW

How ScoutSpark connects.

Open the interactive diagram
High-level ScoutSpark flow: student request, Streamlit interface, preflight, research pipeline, finalization, and results, with model and source inputs.
High-level architecture · ScoutSpark

RUN IT ON YOUR COMPUTER

Bring your own
provider setup.

For live research, install ScoutSpark locally and add the search and model credentials you choose to use. Review budgets before starting a run.

QUICK STARTgit clone https://github.com/FaniBhushan/SparkScout.git
cd SparkScout
python -m pip install -r requirements.txt
python -m streamlit run streamlit_app.py
Full setup instructions