Question 1

What is the difference between RCT Labs and a bare LLM API?

Accepted Answer

A bare LLM API (OpenAI, Claude, Gemini) gives you raw model access with no governance, no persistent memory, and no compliance guarantees. You are responsible for all safety, audit, and PDPA compliance yourself. RCT Labs is a constitutional AI operating system that adds the FDIA framework (deterministic safety), RCTDB (PDPA-compliant persistent memory), JITNA Protocol (agent orchestration), and HexaCore routing (cost optimization) on top of any LLM.

Question 2

Does RCT Labs replace the LLM API or sit on top of it?

Accepted Answer

RCT Labs sits on top of multiple LLM APIs. The HexaCore router includes 7 models (Claude Opus, Kimi K2.5, MiniMax, Gemini Flash, Grok, DeepSeek, and Typhoon v2 for Thai). RCT Labs intelligently routes each query to the appropriate model โ€” you do not need to manage individual API calls. This produces a 3.74x cost reduction compared to always using a premium model.

Question 3

How does RCT Labs achieve PDPA compliance that bare LLM APIs cannot?

Accepted Answer

Bare LLM APIs are stateless โ€” they process a request and forget it. PDPA compliance requires persistence: the ability to prove what data was used, when, and why (Section 33), and the ability to permanently delete a person's data on request (Section 34). RCTDB stores every query as an 8-dimensional record with full provenance. When a data subject requests erasure, their subject_uuid is tombstoned โ€” ensuring no retrievable data remains.

Question 4

What is the 3.74x cost reduction from HexaCore routing?

Accepted Answer

Instead of routing every query to Claude Opus (the most expensive model), HexaCore analyzes each task and selects the optimal model. Simple retrieval tasks go to faster, cheaper models. Complex synthesis tasks go to premium models only when required. Over a 10,000-query enterprise workload, this produces a 3.74x average cost reduction versus always using Claude Opus.

Question 5

Is vendor lock-in a concern with RCT Labs?

Accepted Answer

No. RCT Labs is vendor-neutral by design. The HexaCore router abstracts all 7 models behind a unified interface. If you want to swap Claude Opus for a new model, you update the router configuration โ€” no application code changes required. This is in contrast to building directly on a single LLM API, where switching providers requires significant refactoring.

Feature	Bare LLM API	RCT Ecosystem
Hallucination prevention (deterministic)
PDPA/GDPR audit trail (auto-generated)
Multi-model consensus verification
Persistent memory across sessions
Warm recall (<50ms)
Constitutional kill switch (A=0 โ’ F=0)
Thai-language model support (Typhoon v2)
Vendor-neutral (swap models freely)
Cost optimization (3.74x vs single model)
ISO/PDPA compliance documentation
Direct API access (no middleware)
Open source components (Apache 2.0)
Custom fine-tuning support
Yes Partial No

Scenario	Bare LLM API	RCT Labs	Reason
Enterprise regulated AI (PDPA, finance, healthcare)	โ	โ…	Requires audit trail, right to erasure, and Section 33 explainability โ€” all provided by RCT's RCTDB + JITNA.
Quick prototyping / developer experiments	โ…	โ ๏ธ	Raw LLM API is simpler for one-off experiments. RCT Ecosystem is designed for production-grade deployments.
Agentic workflows with multiple AI agents	โ	โ…	JITNA Protocol provides standardized agent negotiation. LLM APIs have no concept of multi-agent orchestration.
Thai enterprise AI deployment	โ ๏ธ	โ…	HexaCore includes Typhoon v2 (Thai model). PDPA compliance is built-in. RCT Labs is Thailand-native.
Research / academic use	โ…	โ…	Both work. RCT adds reproducibility via signed, verifiable outputs (SignedAI + RCTDB).

RCT Labs vs Bare LLM APIs

Bare LLM APIs (OpenAI, Claude, Gemini, etc.)

RCT Labs Constitutional AI Ecosystem

Feature Comparison Matrix

When to Use Which

Explore the Architecture