SketchWise
SketchWise is specifically designed for mid-sized architectural firms and independent structural engineering contractors who operate with lean teams and tight deadlines. Today, these professionals are forced to spend dozens of billable hours manually translating 2D floor plans into comprehensive structural cost estimates. This tedious, analog workflow creates massive bottlenecks because any minor design revision from a client requires a complete, time-consuming recalculation of the entire project. Furthermore, boutique firms often lack the massive enterprise budgets required to hire dedicated sustainability consultants or invest in extremely complex, high-end parametric simulation software. As a result, they heavily rely on outdated, standard material choices—like traditional concrete and steel—missing out on highly optimized, modern alternatives that could save significant money and reduce carbon footprints. The iterative feedback loop between the lead architect, the structural engineer, and the end client is currently disjointed and incredibly slow. Clients frequently grow frustrated because they cannot receive rapid, accurate visualizations of how a design tweak will impact their final construction budget. Without an automated way to verify load distributions and compliance, human error inevitably creeps into the manual Bill of Materials, leading to costly on-site material shortages or wasteful over- ordering. Engineers are also overwhelmed by the sheer volume of mundane data entry required to cross-reference local material costs with modern building codes. Ultimately, SketchWise eliminates this friction by instantly transforming basic uploads into optimized, compliant, and cost-saving blueprints, empowering smaller firms to punch above their weight and compete with large enterprise agencies.
Preliminary score
- Value for the user15.5 / 25
A specific problem and a clear outcome from solving it.
- Prototype and use of AI16.0 / 30
A working core scenario, and what AI actually contributes to the result.
- Quality testing9.0 / 20
Tests, examples of failures, and comparison with the current approach.
- Feasibility4.0 / 15
Data requirements, running costs and a clear next step.
- Originality5.0 / 10
A meaningful difference from common solutions.
What the judges wrote
- Claude Opus 5.5Preliminary44/100
- Strengths
- Your target user is clearly named: mid-sized architecture firms and independent structural contractors who re-estimate costs by hand after every design revision. The README describes a coherent full-stack build (React/Three.js front end, Express/SQLite back end, file upload to AI-extracted variables to original-vs-optimized cost). Your testing honestly reports concrete software defects, including the paywall leak through client state, the 22px button collapse and the silent 404s after session loss, and explains how you fixed each one.
- To improve
- Your testing covers the app shell but not the AI. Run a set of real floor plans through SketchWise and compare the extracted dimensions, quantities and costs against a manual or spreadsheet estimate, then report the error rates and the plans it misread. Claims of structural optimization and code compliance 'without compromising structural integrity' are strong for an LLM, so show how they are checked or narrow the scope to cost estimation. Add the missing feasibility pieces: where material prices and building codes come from, the per-analysis model cost, and a concrete pilot step. Also remove the live API key and default admin password from your public submission.
- GPT-6 AstraPreliminary55/100
- Strengths
- Your focus on lean architectural and engineering teams gives SketchWise a concrete workflow: revising material estimates when plans change. Your documented architecture gives AI a substantive role in extracting project variables and suggesting cost-saving alternatives. Your testing write-up identifies specific paywall, interface, and session failures with corresponding fixes, rather than offering only generic reliability claims.
- To improve
- Your next validation should show annotated plans processed end to end, comparing extracted dimensions, quantities, costs, and turnaround time against an engineer's spreadsheet baseline, including failure cases. Your structural-integrity and compliance claims need engineering validation and explicit human approval before outputs can be relied on. Your feasibility case should identify local pricing and building-code sources, inference cost per plan, and a concrete pilot milestone. Your deployment should also replace the shared default admin password and rotate any real API credential included in the submission.
The team
Ramiz Samadov
business
Tofig Valizade
product
Nurlan BabayevLead
developer