This folder is for comparative notes, external landscape analysis, and longer-form research that should not live in the main product, benchmark, or roadmap docs.
Current documents:
- solver_safety_and_validation_hardening.md Solver safety architecture for explicit failure modes, input validation, convergence safeguards, post-solve verification, structured diagnostics, solver-run artifacts, and frontend mutation guards.
- automation_gaps.md Full automation gap analysis for Stabileo/Dedaliano: what is already automated, what engineers still do manually (high and medium impact), first-mover opportunities no competitor has, an automation maturity ladder, a prioritized implementation matrix, and a competitor coverage map.
- competitive_displacement_by_step.md Which roadmap steps displace which competitors, what each step removes from the incumbent stack, and the likely savings/positioning impact.
- structural_engineering_software_market_research.md Competitor pricing and market-position research across CSI, Dlubal, Bentley, CYPE, SkyCiv, ClearCalcs, and other structural software vendors.
- cypecad_competitive_gap_and_parity_plan.md Unified CYPECAD note: corrected gap framing plus phased parity plan, with emphasis on RC automation, BBS, reports, and the narrower remaining genuine solver gaps.
- rc_design_and_bbs.md Why RC design, reinforcement schedules, and later BBS generation should be explicit priorities, with solver vs product responsibilities and staged delivery order.
- post_roadmap_software_stack.md Best software products to build once the core roadmap lands: RC design/BBS studio, report OS, QA/review assistant, firm workspace, parametric configurators, interoperability, cloud comparison, and education.
- ai_provider_architecture.md Recommended AI integration architecture: frontend calls an internal capability-based AI layer, which routes to Claude/OpenAI/Kimi/local/future providers through backend adapters.
- open_source_vs_hosted_ai_boundary.md Recommended split between open-source AI surfaces and hosted/private AI moat: keep baseline AI usefulness in OSS, keep orchestration, scale, collaboration, and premium automation in the paid layer.
- open_source_solver_comparison.md Comparison of Dedaliano against major open-source structural / FEA solver projects.
- numerical_methods_gap_analysis.md Large-model numerical-methods gap analysis and corrected performance priorities.
- competitor_element_families.md Competitor shell/element-family matrix and the highest-value remaining gaps.
- shell_family_selection.md Shell-family selection rules, defaults, and product guidance.
- beyond_roadmap_opportunities.md Research-backed opportunities beyond the solver roadmap: code checking, ML surrogates, FEMA P-58, topology optimization, BIM-IFC, CLT, seismic automation, generative design, and a nuanced WebGPU assessment (GPU helps visualization/postprocessing/topology, but sparse direct stays on CPU).
- webgpu_solver_renderer_analysis.md WebGPU fit analysis for the renderer vs solver, with ROI and sequencing guidance.
- cholmod_nvidia_gpu_research.md CHOLMOD + NVIDIA GPU feasibility research for Dedaliano: architecture fit, realistic speed expectations, remote-server implications, and benchmark plan.
- lean_formal_verification.md Lean formal-verification research plan for the solver core, with ROI-ranked proof targets, theorem ladder, and phased implementation strategy.