ConductColony Intelligence

The first colony platform built on real statistical ML.

Anomaly detection, yield forecasting, and an LLM assistant that runs on your colony's own data. No rule-based alerts dressed up as AI.

Try on your colony See the methods

Intelligence Layer

real-time · audit-trailed

Four capabilities, one platform

Each capability surfaces inside the modules you already use. Animals, breedings, litters, and dashboard. No separate app, no extra logins.

Litter L-2026-044 · Anomaly

JO_CatSperD · χ²(2) = 6.84

Genotype distribution

ExpectedObservedFlag

⚠ Possible embryonic lethality

KO1/KO1 observed 0/12, expected 3.0. p = 0.042 · BF = 8.2

χ²

6.84

Capability 01

Anomaly Detector

χ² and Bayesian flagging on every litter. Spot embryonic lethality, sex-linked transmission, and CRISPR mosaicism the moment a litter genotypes out of expectation.

Cross Search · Goal: Cre+;flox/flox

JO_CatSperD · 12 mice by 2026-07-01

Colony graph · 247 candidate pairs

Top 3 plans

#1 M-187 × F-204

ETA 2026-06-2214.2 pups

#2 M-211 × F-198

ETA 2026-06-2911.6 pups

#3 M-156 × F-219

ETA 2026-07-049.8 pups

Capability 02

Cross Recommender

Tell us your goal genotype and target date. The recommender searches your colony for the best ranked pair plans, scored on yield, inbreeding, time-to-target, and cost.

Forecast · 6mo · P-19-0042

80 Cre+ pups · target 2026-09-30

Projected cage demand

BreedingLittersHolding

Breeders needed

Peak cages

24 /32

Per-diem

$3,400

Capability 03

Production Planner

Reverse-plan from an IACUC target to a breeder count, projected cage demand, and per-diem cost. Aware of your purchased rack capacity and protocol-approved animal limits.

Linnaeus · Read-only

schema: prisma · 14 tables

JDYou

Which breeding pairs have not produced a litter in the last 60 days, sorted by strain rarity?

Generated SQL · review before runSELECT only

SELECT bp.id, bp.maleId, bp.femaleId,

s.name AS strain,

s.animals_count, MAX(l.birthDate) AS last_litter

FROM breeding_pairs bp

JOIN strains s ON s.id = bp.strainId

LEFT JOIN litters l ON l.breedingPairId = bp.id

WHERE bp.status = 'active'

GROUP BY bp.id

HAVING last_litter IS NULL

OR last_litter < NOW() - INTERVAL '60 days'

ORDER BY s.animals_count ASC

3 results · click any row to open pair

Pair	Strain	Rarity	Last litter
B-32	JO_CatSperD	4 animals	never
B-19	JO_Efcab9	7 animals	2026-02-14
B-44	JO_WT	18 animals	2026-02-28

Capability 04

Linnaeus Assistant

Ask your colony anything in plain English. Linnaeus translates the question into SQL, shows you the query for review, and returns the answer in seconds.

Why this is different

Most colony software vendors describe rule-based alerts as AI. ConductColony Intelligence uses real statistical methods, transparent math, and a real LLM with auditable output.

Statistical method

χ² goodness-of-fit, Bayesian posterior, Cox proportional hazards, gradient-boosted regression.

Trained on your data

Models use your colony's own breeding history. No third-party training corpus, no shared models across organizations.

Math you can audit

Every recommendation comes with the underlying numbers. p-values, confidence intervals, and full computation traces.

LLM with guardrails

Linnaeus runs against a pinned schema with a human-readable SQL preview before execution.

All Intelligence features included on Pro

$99 per month. No upsell tier, no metered usage, no per-seat fee. Upgrade when your colony outgrows the Lab plan.

See full pricing Live demo

Bring statistical rigor to your breeding decisions

Start free, upgrade to Pro when you want Intelligence on your colony.

Start Free