Same drug. Brand, generic, abbreviation. One canonical record.
Tylenol Extra Strength 500mg shows up as "Acetaminophen 500mg Caplets" in one system and "APAP 500 MG TAB #100" in another. Across formularies, pharmacy systems, and claims data, the same medication appears dozens of ways. Fuzzy resolves them all — automatically.
Brand vs generic naming is inconsistent. “Tylenol”, “Acetaminophen”, and “APAP” are the same active ingredient — systems don’t agree.
NDC codes vary across sources. Different packagers, reformulations, and lot sizes create duplicate entries for identical drugs.
Strength and count formats diverge. “500mg x 100”, “500 MG TAB #100”, “0.5g/100ct” — same product, different notation.
Dirty drug data affects patient safety. Duplicate formulary entries, billing errors, and reconciliation failures have real consequences.
Why Fuzzy
Pharma-aware normalization. Maps brand names to active ingredients, understands dosage form equivalence and strength conversions.
Self-improving. Every pharmacist correction trains the system. Formulary reconciliation gets faster with each run.
Fully explainable. See exactly why every match was made — critical for compliance, audit trails, and regulatory review.
Strength/count veto fields. Auto-reject when dosage or count mismatches — “500mg” never matches “250mg”.
Deeply configurable. Tune drug matching rules, dosage normalization, veto fields, and confidence thresholds — every formulary is different.
Four-Tier Matching Pipeline
T1
Exact Match — Normalized
Free
T2
Fuzzy Consensus — 3/5 vote
Free
T3
LLM Reasoning — Ambiguous pairs
~$0.001
T4
Human Review — Feedback loop
You
80% free15% LLM5% human
LLM + Human Review
The LLM maps brand names to active ingredients and understands dosage form equivalence — “APAP” = “Acetaminophen” = “Tylenol”. It reasons about therapeutic class, strength, and packaging to resolve ambiguous pairs. Human pharmacists review edge cases, and every decision is permanently cached for audit.