Versus LangSmith / observability tools
They're engineer-facing tracing for LangChain (mostly). We're enterprise governance + cross-provider + customer-facing receipts. Different audiences; LangSmith won't become us, we won't become them.
Versus AI memory tools (Mem0, Letta)
They're memory storage for stateful agents. We're audit + portability + governance. Complementary; many teams use both.
Versus enterprise AI governance products
Most enterprise AI governance products focus on policy enforcement upfront (block X, allow Y). We focus on attribution + audit + portability after the fact. Policy enforcement misses novel attacks; attribution catches them post-hoc. Different philosophies; we think attribution wins at scale.
Versus competitive vendors (SIGIL, Vinsta, MerchantGuard, etc.)
We lead on surface area (more AI tools integrated, more frameworks). They lead on specific niches (e.g. MerchantGuard's payment integration). The competitive landscape is fragmented; we're betting on the metalayer position.
Versus Anthropic / OpenAI / Google native features
They'll ship vendor-internal features (Claude Projects, ChatGPT Memory). They won't ship cross-provider portability — it hurts their lock-in. Cross-provider has to live outside any single vendor; that's us.
Why all four matter together
Cross-provider without cryptographic receipts = inadequate audit. Cryptographic receipts without open spec = vendor lock-in. Open spec without buyer-side trust = no marketplace value. Buyer-side trust without cross-provider = no portability moat. The combination is what makes the position defensible.