Triplets is an Enigma Vault product
How it works

You can't prompt
your way to the truth.

Every team tries the same four fixes before they call us. Here's why each one fails — and what it actually takes to solve this.

The four fixes that don't work

Four ways to ask a model
to behave. None of them hold.

“We'll just tell it to never hallucinate.”

A prompt is a request, not a constraint. The model has no ground truth to check itself against, and instruction-following is probabilistic — it slips on edge cases, long contexts, and every model update. Hallucination is the generation mechanism working as designed. It is not the model disobeying you.

“We'll tell it to refuse when it's unsure.”

Models are badly calibrated. Stated confidence doesn't track correctness, so the model can't reliably know what it doesn't know. Tune it toward caution and it refuses questions your data can answer. Tune it toward coverage and the fabrication comes back. You're turning a dial you can't see.

“We'll fine-tune it on our domain.”

Training lowers the error rate. It can't set a floor of zero. And behavior baked into weights can't cite where an answer came from, doesn't update when your data changes tomorrow, and regresses every time you swap or upgrade the model.

“We'll set temperature to zero.”

Deterministic decoding makes the output repeatable, not correct. A zero-temperature model gives you the same wrong answer every time, with the same fluent confidence. Consistency and accuracy are different things.

The point

Hallucination is an architecture problem, not an obedience problem. Generation cannot police itself. Correctness has to be enforced by a system that sits outside the model and knows what is verified, what is not, and where the line is.

A hierarchy of guarantees

Every answer has to earn
the right to exist.

Most RAG retrieves some chunks and hopes the model behaves. Triplets treats answering as a hierarchy of guarantees, and only generates when generation is the right tool.

L1

Deterministic lookup

Strongest guarantee

Exact facts answered from structured, verified values. No generation, no room to improvise — used whenever the data allows it.

L2

Relationship traversal

Verified paths

Questions that span entities and documents are answered by walking verified relationships: which policy covers which clause, which drug interacts with which order.

L3

Retrieval + reasoning

Constrained generation

Open questions go to constrained generation. The model reasons over retrieved evidence but stays bound to it. Verified facts override the model — never the reverse.

Past the boundary, it refuses

When none of the three layers can answer from verified data, Triplets declines, tells the user why, and logs the gap. A refusal is recoverable. A fabrication is not.

Self-healing

It gets more accurate
the more it is used.

Most systems throw failures away. Triplets treats every failure as fuel.

01 · DETECT

Detect

Every refusal, partial answer, and failed eval is captured as a signal — not noise.

02 · CLASSIFY

Classify

The failure is diagnosed by type: missing data, broken relationship, conflicting sources, or retrieval miss.

03 · REPAIR

Repair

The knowledge base is corrected at the source, so the same class of question answers correctly next time.

04 · REMEMBER

Remember

A memory ledger records what failed, what was fixed, and why. Accuracy compounds instead of resetting.

Substrate-agnostic, not content-agnostic

Add Triplets without
rebuilding your stack.

Triplets re-ingests your corpus and models it into its own structured stores of entities, relationships, and verified values. Then it governs generation at query time using that model.

That makes it substrate-agnostic: it runs alongside whatever vector database, graph, or pipeline you already have. Your stack stays where it is — Triplets governs how its output becomes an answer.

What it is not is content-agnostic. It's built for entity-rich, fact-verifiable data — and it's honest about where that ends.

Your app / product query
Triplets — reliability layer governs answers
Entities · Relationships · Verified values Triplets stores
Vector DB · Graph · Pipeline your stack, unchanged

re-ingest once · govern at query time

The report is the pitch

See it run on your own data,
in 48 hours.

Bring the questions your AI gets wrong today. We run them against your stack and against Triplets, side by side, and hand you the before-and-after report. No integration work. No commitment.