Build the ration, see predicted milk against the model your practice trusts, dial toward a milk target, and open RIQ-Core™ v1.3 when you need the full rumen story — not just a spreadsheet endpoint.
Graieva puts predicted milk at the center — simple enough for the dairy floor, deep enough for the consulting nutritionist, with feed import, prices, saved sessions, and practice overrides instead of another spreadsheet.
NASEM, CNCPS, Blend, RIQ-Core, or Blend + RIQ — choose the basis you trust, without juggling separate engines.
Capture the small deviations your nutritionists already use. Name the profile, tune the knobs, and select it like any other base.
Set the milk you need and rebalance. Cost-aware suggestions, multi-pen runs, and push updates into recipes and pens when you’re ready.
Balance day-to-day in the language your farm already uses, then open the full mechanistic rumen story for stress, SARA risk, and efficiency.
Graieva leads with dairy ration balancing, then opens into feed intelligence, herd analytics, and API access — all driven by RIQ-Core™ v1.3.
CNCPS, NASEM, Blend, RIQ, and practice profiles — predicted milk first, hit targets, update pens and recipes.
Learn moreProprietary feed library, recipe builder, Biological Twin cold-start, and cache audit.
Portfolio audits, alerts, uncertainty, sensitivity, and engine regression tooling.
Programmatic access to mechanistic simulation, benchmarks, and tenant-scoped data.
Learn moreDairymen and nutritionists deserve a straight answer. RIQ-Core™ is Graieva’s proprietary ruminant intelligence engine: a full-tract mechanistic solver we built to capture the dynamics that static CNCPS-style tables leave on the table. It is our model — transparent enough to audit, powerful enough to couple into ration balancing — not a borrowed logo on old math.
Coupled differential equations on a 6-minute step for 300 hours — rumen N, microbial biomass, protozoa, and ammonia evolving together, not a single lookup row.
We designed RIQ-Core™ as a newer, more open alternative to factor-table engines: dynamic pools, passage that moves with DMI/BW, and outputs you can audit.
When you open the full rumen story from balancing — or dig into alerts, flux, and uncertainty — this is the engine behind those answers: mass-balance checked, versioned, and comparable to legacy baselines.
See the equations, residuals, and side-by-side checks against CNCPS/NASEM. You should understand why a number moved — not just that it did.
We are not CNCPS v6.5, not a CNCPS license, and not a skin over someone else’s proprietary factor model. Comparisons exist so you can judge agreement and divergence honestly.
RIQ-Core is deterministic biology math. Learning may help cold-start feed kinetics elsewhere; the core solver is mechanistic.
Daily balancing still lets you predict against CNCPS, NASEM, Blend, or your practice’s own model. RIQ is there when the rumen’s path — not only the endpoint — matters.
Weather, feed chemistry, and animal inputs still drive outcomes. RIQ-Core makes the biology visible so you can adjust the diet with eyes open.
Think of CNCPS/NASEM as the languages many farms already balance in. Think of RIQ-Core™ as the living rumen model underneath Graieva — integrating intake, microbes, urea recycle, VFA/pH, methane, and N fate over a 300-hour horizon so you can see stress, inefficiency, and opportunity that a static sheet will miss.
Most days you pick CNCPS, NASEM, Blend, or a practice profile and aim at a milk target. When you want the living rumen story — heat load, microbial protein, SARA risk, methane — you open RIQ-Core™. Same farm. Clearer biology.
RIQ-Core™ v1.3 integrates four coupled differential equations on a fine time step across hundreds of hours, capturing dynamics that static factor models miss: urea recycling under low rumen ammonia, protozoal predation of bacteria, and passage rates that move with intake and body weight.
dR/dt — rumen N + urea recycle − microbial uptake − passagedB/dt — microbial biomass − protozoal predation − passagedP/dt — protozoal pool with predation efficiency εdNH₃/dt — deamination + urea − uptake − absorptionOver 300 hours, these loops converge to a steady state. RIQ-Core™ v1.3 captures that state — and the path to get there — in under a second per pen.
Start with day-to-day balancing. Open the deeper rumen tools when you need uncertainty, sensitivity, or herd-wide economics.
Predict milk against CNCPS, NASEM, Blend, RIQ, or your practice profile. Aim at a milk target and push updates to pens.
SARA exposure, low N efficiency, urinary-N ammonia, methane intensity, WEV confidence and mass-balance residuals — with action recommendations.
VFA mass balance with saliva + ration buffer capacity. SARA flag triggers when pH spends > 3 h/day below 5.8.
Tier-2 methane that shifts with acetate:propionate, plus mechanistic fecal vs urinary N partition for sustainability scoring.
Sample feed chemistry from lab CVs and get P10 / P50 / P90 envelopes on the 300-hour traces — see when a recommendation is robust.
Perturb each kinetic parameter ±10% and rank impact on MCP flow, N efficiency, and predicted methane. Know which inputs matter.
Side-by-side checks so you can see where RIQ-Core™ agrees with legacy baselines — and where dynamic biology diverges.
pgvector cold-start: embed a new feed and pull blended kinetics from the anonymized Global Commons. No more guessing kd/kp.
Rank every pen across every client by economic loss in one click. Tagged findings, herd rollups, and exportable narratives.
Your proprietary feed library never leaves your tenant. Anonymized embeddings power the Global Commons, but chemistry stays yours.
Scale → genetic prior (Bayesian shrinkage) → WEV inverse solver. Confidence < 0.6 raises a 'go check the pens' alert.
Postgres row-level security with a SECURITY DEFINER access helper. No cross-tenant leakage; benchmarks enforce a k-anonymity floor.
Diet × kinetics × version → cached steady state. Kinetics edits invalidate just the 276–300 h window, not the whole trace.
Equation registry, mass-balance report, and 300-hour CSV trace — one click from any pen.
Tier license for access. Flat $0.007 per milking cow per day — no volume ladder. Herd for a single site, Studio for consultancies.
14 days · 1 user · full engine access
license/mo · + $0.007/cow/day · 3 users
Click card to estimate with this plan
Get startedlicense/mo · + $0.007/cow/day · 5 users · up to 25 workspaces
Click card to estimate with this plan
Open dashboardCo-ops · integrators · genetics
Select Herd or Studio above (or here). License + $0.007/cow/day · 30-day month
RIQ-Core™ runs a 300-hour mechanistic integrator — not a static CNCPS shell. If factor tables are enough, legacy tools are fine.
Embed the mechanistic engine in your own product. Priced separately from Herd/Studio SaaS. View full API overview →
RIQ-Core™ is Graieva’s own living rumen model — tracking nitrogen, microbes, protozoa, acidity, methane, and manure nitrogen over time. It was built as a newer, more auditable alternative to static CNCPS-style tables. It is not CNCPS itself, not a CNCPS license, and not a black-box milk AI. Day-to-day balancing can still use CNCPS or NASEM; RIQ is the deeper biology when you need it.
Yes. Choose NASEM, CNCPS, Blend (leans conservative when the two disagree), RIQ-Core, Blend + RIQ, or a profile your practice owns. One choice drives the milk prediction.
No. Day-to-day balancing speaks CNCPS and NASEM fluently, but RIQ-Core™ v1.3 is our own mechanistic engine — urea recycling, protozoal predation, and dynamic passage included. Side-by-side comparisons are there when you want to audit the science.
Monte Carlo mode samples feed chemistry from your lab CVs and returns P10 / P50 / P90 envelopes on every 300-hour trace. Tornado mode ranks ±10% perturbations of every kinetic parameter against MCP, N efficiency, and CH₄.
The Biological Twin RPC searches a pgvector index of anonymized global feeds and returns blended kinetics weighted by similarity. Your proprietary chemistry is never exposed to other tenants.
Sign in, pick CNCPS, NASEM, Blend, or your practice profile, and aim at a milk target — with RIQ-Core™ ready when you need the deeper rumen story.