Budget Allocation

A lean, pragmatic distribution across compute, data, tooling, personnel, and contingency — built to maximize academic impact per dollar.

Funding Deadline: All allocated funds must be committed before the end of June 2026.
Lean by Design

Initiated with under $15K USD

HuskyBot is being built on a fraction of what comparable enterprise AI efforts cost — thanks to open-weight models, free academic corpora, and parameter-efficient fine-tuning. The breakdown below shows how each share is allocated by category.

Allocation by Category

Loading chart...

Detailed Breakdown

Compute & Training

40%

Cloud GPU rental for model fine-tuning, training runs, hyperparameter tuning, and evaluation benchmarks.

GPU cloud instancesTraining compute hoursModel evaluation runs

Data & APIs

20%

API access for academic databases, data processing tools, and vector database hosting for the RAG system.

Semantic Scholar APIVector databaseData processing tools

Software & Tools

17%

Development tools, hosting services, monitoring solutions, and web interface deployment infrastructure.

Development environmentWeb hosting & deploymentMonitoring & logging

Personnel & Training

13%

Student assistant hours, training materials, documentation development, and user-guide creation.

Student research assistantsDocumentation & guidesTraining workshops

Contingency

10%

Reserved share for unexpected costs, additional compute needs, or scope adjustments during development.

Emergency compute needsScope adjustmentsRisk mitigation