Protect software IP by automatically 'sloppifying' code with AI-driven obfuscation

Software vendors, SaaS companies, mobile and edge app builders, and teams shipping WASM or client-side ML increasingly face IP and model-theft risk as stronger decompilers and model-extraction tools driven by AI make reverse engineering easier. This is a measurable commercial problem: the addressable market is roughly $6.0B (500,000 software‑centric organizations at about $12K ACV), with a market score of 92/100 and revenue potential 88/100, and competition that is currently medium but likely to intensify. You could build an automated, build‑time obfuscation platform that "sloppifies" source or compiled artifacts with AI-driven, semantics‑preserving transformations tuned per target (JavaScript/WASM, mobile native, Python bytecode, and model weights), delivered as CI/CD plugins with deterministic mapping files for debugging. Core capabilities would include configurable protection levels, measurable resilience scores via automated attack emulation, continuous transformation updates against new decompilers and model-extraction techniques, and lightweight SDKs to keep runtime and size overhead low. A go‑to‑market approach might price per-repo or per-build subscription with optional professional services to harden high-value modules, aiming at the ~$12K ACV mid-market bracket. Market timing is favorable: AI-based reverse engineering, growing client-side compute, and the shift‑left security trend create strong demand for build-integrated protections that scale with existing pipelines. To win, be candid about limits—obfuscation is an arms race and not a silver bullet—but differentiate through tight CI/CD integration, repeatable resilience metrics, minimal performance impact, and an operations model that continuously adapts transformations in response to attacker tooling.