← pcfic.com
COMPASS
LIVE DEMO
Compass
Pacific Platform · Astrognosy AI

Compass
Route Smarter. Pay Less.

Every query is scored for confidence against a growing corpus. High-confidence queries resolve on CPU. Everything else routes to GPU. The corpus builds itself — your cost floor drops automatically over time.

$0.0003
CPU path/query
$0.01
GPU path/query
<2ms
Routing decision
33×
Cost ratio
Input
Query
Coverage
PCF Engine
confidence ≥ 0.70 · serve
CPU · $0.0003
confidence < 0.70 · infer
GPU · $0.01
How It Works
Step 1
Confidence Scoring
The query is scored against the current corpus. High score = the answer is already well-covered. Low score = the system isn't confident enough to answer alone.
Step 2
Route Decision
Confidence ≥ 0.70 → CPU path. Served locally in under 2ms at a fraction of GPU cost. Below threshold → routes to GPU for accurate inference.
Step 3
CPU Path
Answer served directly from the corpus. Deterministic, auditable, and cached — repeat queries cost nothing. No LLM call made.
Step 4
Corpus Grows
GPU responses feed back into the corpus automatically. The next similar query routes to CPU. Cost floor drops over time without any manual labeling.
Threshold Calibration · Empirical Benchmark

Routing precision and local-serve economics vary with the confidence threshold. Measured on a 48-claim benchmark spanning science, history, technology, medical, and economics domains — including adversarial cases designed to trap keyword classifiers.

Threshold Local Serve% routed to CPU Routing Precisionaccuracy when CPU serves REFUTED Recallfalse claims correctly flagged GPU Savingsvs. all-GPU baseline
0.50 72.9% 68.6% 45.0% 79%
0.55 72.9% 68.6% 45.0% 79%
0.60 64.6% 71.0% 55.0% 75%
0.65 50.0% 70.8% 65.0% 65%
0.70 deployed 25.0% 75.0% 85.0% 46%

Why 0.70? Lower thresholds serve more queries locally but increase false-positive CPU routes — claims incorrectly marked as supported. At 0.70, routing precision reaches 75% with 85% REFUTED recall: the minimum acceptable bar for production use. As the corpus grows via the response flywheel, more claims naturally clear 0.70 — local serve rate and savings increase without touching the threshold.

Verify Claim
Detect Contradiction
Claim to Verify
Reference Documents (3)

Detects structural divergence between two passages — works best on texts with meaningfully different vocabulary and flow, not just different numbers in the same sentence structure.

Text A
Text B