Infermap MCP Server

Moved. This repo has moved into the benzsevern/goldenmatch monorepo at packages/python/infermap (and packages/typescript/infermap)/. This repo is archived; new development happens in the monorepo.

---

infermap is a schema-mapping engine. Give it any two field collections (CSVs, DataFrames, database tables, in-memory records) and it figures out which source field corresponds to which target field, with confidence scores and human-readable reasoning. Available as a Python package on PyPI and a TypeScript package on npm, with mapping decisions verified bit-for-bit by a shared golden-test parity suite.

Table of contents

• Install
• Quick start
• How it works
• Features
• Which package should I use?
• Custom scorers
• CLI examples
• Config reference
• Documentation
• License

Install

Python

pip install infermap

Optional database extras:

pip install infermap[postgres]   # psycopg2-binary
pip install infermap[mysql]      # mysql-connector-python
pip install infermap[duckdb]     # duckdb
pip install infermap[all]        # all extras

TypeScript / Next.js

npm install infermap

Zero runtime dependencies in the core entrypoint. Compatible with Next.js Server Components, Route Handlers, Server Actions, and the Edge Runtime out of the box. See the package README for the full reference.

Quick start

Python

import infermap

# Map a CRM export CSV to a canonical customer schema
result = infermap.map("crm_export.csv", "canonical_customers.csv")

for m in result.mappings:
    print(f"{m.source} -> {m.target}  ({m.confidence:.0%})")
# fname -> first_name  (97%)
# lname -> last_name   (95%)
# email_addr -> email  (91%)

# Apply mappings to rename DataFrame columns
import polars as pl
df = pl.read_csv("crm_export.csv")
renamed = result.apply(df)

# Save mappings to a reusable config file
result.to_config("my_mapping.yaml")

# Reload later — no re-inference needed
saved = infermap.from_config("my_mapping.yaml")

TypeScript

import { map } from "infermap";

const crm = [
  { fname: "John", lname: "Doe", email_addr: "j@d.co" },
  { fname: "Jane", lname: "Smith", email_addr: "j@s.co" },
];

const canonical = [
  { first_name: "", last_name: "", email: "" },
];

const result = map({ records: crm }, { records: canonical });

for (const m of result.mappings) {
  console.log(`${m.source} → ${m.target}  (${m.confidence.toFixed(2)})`);
}
// fname       → first_name  (0.44)
// lname       → last_name   (0.48)
// email_addr  → email       (0.69)

For Next.js, drop it directly into a Route Handler — works on Edge Runtime with zero config:

// app/api/infer/route.ts
import { map } from "infermap";
export const runtime = "edge";

export async function POST(req: Request) {
  const { sourceCsv, targetCsv } = await req.json();
  const result = map({ csvText: sourceCsv }, { csvText: targetCsv });
  return Response.json(result);
}

How it works

Each field pair runs through a pipeline of 7 scorers. Each scorer returns a score in [0.0, 1.0] or abstains (None/null). The engine combines scores via weighted average (requiring at least 2 contributors), then uses the Hungarian algorithm for optimal one-to-one assignment.

The engine also applies common-prefix canonicalization — automatically stripping schema-wide prefixes like prospect_ so that City vs prospect_City is compared as City vs City. And optional confidence calibration transforms raw scores into calibrated probabilities post-assignment (ECE from 0.46 to 0.005 on real-world data).

Read the full architecture →

Features

Full feature parity matrix →

Which package should I use?

What's new in v0.3

+18.3pp F1 on real-world data from four compounding improvements:

v0.2 baseline    F1 0.657
+ min_conf 0.2   F1 0.765  (+10.8pp — empirically tuned threshold)
+ prefix-strip   F1 0.821  (+5.6pp  — City vs prospect_City now works)
+ InitialismScorer F1 0.840 (+1.9pp  — ASSI, CONSC, RELATIT now work)

New features:

• Domain dictionaries — MapEngine(domains=["healthcare"]) loads curated aliases for your domain. Ships: generic (default), healthcare, finance, ecommerce. See examples/09_domain_dictionaries.py.
• Confidence calibration — MapEngine(calibrator=cal) transforms raw scores into calibrated probabilities. Ships: IsotonicCalibrator, PlattCalibrator. Valentine ECE: 0.46 → 0.005. See examples/10_calibration.py.
• InitialismScorer — matches abbreviation-style column names (assay_id ↔ ASSI). ChEMBL F1: 0.524 → 0.819.
• Common-prefix canonicalization — automatically strips prospect_, assays_, etc. before fuzzy matching.
• Valentine corpus — 82 real-world schema-matching cases from the Valentine benchmark for accuracy testing.
• Full TypeScript parity — all new features ported. 186 TS tests. Benchmark F1 within 0.0005 of Python.

Custom scorers

Python

import infermap
from infermap.types import FieldInfo, ScorerResult

@infermap.scorer("prefix_scorer", weight=0.8)
def prefix_scorer(source: FieldInfo, target: FieldInfo) -> ScorerResult | None:
    if source.name[:3].lower() != target.name[:3].lower():
        return None
    return ScorerResult(score=0.85, reasoning=f"Shared prefix '{source.name[:3]}'")

from infermap.engine import MapEngine
from infermap.scorers import default_scorers

engine = MapEngine(scorers=[*default_scorers(), prefix_scorer])

TypeScript

import { MapEngine, defaultScorers, defineScorer, makeScorerResult } from "infermap";

const prefixScorer = defineScorer(
  "prefix_scorer",
  (source, target) => {
    if (source.name.slice(0, 3).toLowerCase() !== target.name.slice(0, 3).toLowerCase()) {
      return null;
    }
    return makeScorerResult(0.85, `Shared prefix '${source.name.slice(0, 3)}'`);
  },
  0.8 // weight
);

const engine = new MapEngine({
  scorers: [...defaultScorers(), prefixScorer],
});

CLI examples

The CLI works the same way in both packages:

# Map two files and print a report
infermap map crm_export.csv canonical_customers.csv

# Map and save the config (Python: --save, TS: -o)
infermap map crm_export.csv canonical_customers.csv -o mapping.json

# Apply a saved mapping to rename columns
infermap apply crm_export.csv --config mapping.json --output renamed.csv

# Inspect the schema of a file or DB table
infermap inspect crm_export.csv
infermap inspect "sqlite:///mydb.db" --table customers

# Validate a saved config against a source
infermap validate crm_export.csv --config mapping.json --required email,id --strict

Config reference

Both packages accept an engine config (scorer weight overrides + alias extensions). Python uses YAML, TypeScript uses JSON; the shape is identical.

# Python: infermap.yaml
domains:
  - healthcare
  - finance
scorers:
  LLMScorer:
    enabled: false
  FuzzyNameScorer:
    weight: 0.3
aliases:
  order_id:
    - order_num
    - ord_no

// TypeScript: infermap.config.json
{
  "scorers": {
    "LLMScorer":       { "enabled": false },
    "FuzzyNameScorer": { "weight": 0.3 }
  },
  "aliases": {
    "order_id": ["order_num", "ord_no"]
  }
}

See infermap.yaml.example for a full annotated reference.

Documentation

• 📖 Wiki — full reference for both languages
  • Getting Started
  • Python API
  • TypeScript API
  • Python vs TypeScript — migration guide
  • Scorers
  • Architecture
  • FAQ
• 🌐 Documentation site
• 🧪 Examples
  • Python examples — 10 numbered scripts covering basic mapping, databases, custom scorers, config, domain dictionaries, calibration, and score-matrix introspection
  • TypeScript examples — basic mapping, Next.js Edge Runtime, custom scorer, databases, domain dictionaries, save/reuse
• 📓 Open in Colab — Python notebook
• 💬 GitHub Discussions
• 🐛 Issue tracker

Author

Ben Severn

License

MIT