OB
Omniblend Discover · Blend Quality & Cost Savings
—
—
Discover
Lubricant Master
Recipe Workbench
Savings Ledger
Analytics
Quality Findings
OMNIBLEND
Stop Guessing. Start Gauging.
Omniblend Discover eliminates the "Blind Period" between formulation and results, turning real-time quality control into verifiable fiscal gain.
Quality you can trust, in dollars you can trace.
The "Blind Period" — where money leaks
Traditional blending operates without molecular visibility
  • Theoretical recipes meet real receipts. Plants blend on the textbook formula even when the additive batch arrives 5% over-potent or a base oil viscosity drifts 6% off standard.
  • Variance discovered only post-blend. The lab test happens after the kettle is full — by then a corrective top-up is the only way to recover spec.
  • Each correction burns additive cost. Adding a 0.5% top-up of a $2,500/MT additive to replace what should have been $850/MT base oil is a $1,650/MT penalty per top-up event.
  • Tribal calculations vary by shift. Senior blender judgment carries the 90% RFT — a junior on the same line drops it.
The Discover response · Blend QA before the valves open
Capture · Predict · Rebalance · Receipt
  • Capture — digitise the COA the moment it arrives. Per-component KV, viscosity, Ca, Zn, P actuals captured against standard.
  • Predict — mass-balance + perturbation model computes the finished blend KV and elemental fingerprint before any material moves.
  • Rebalance — hold the licensed-chemistry compliance floor; swing the heavy/light base oil pair to land target KV. Multi-variant solve handles physical, chemical and elemental targets simultaneously.
  • Receipt — every avoided correction recorded with its batch, the variance that triggered it, and the $ saved. Shown to be arrived at, never asserted.
Key Value Pillars
Three pillars that make quality assurance verifiable and financially traceable.
🔬
Real-Time Quality Assurance
AI-driven predictive modeling monitors blend integrity as it happens, correcting deviations before they reach the lab.
📒
The Savings Ledger
Every intervention is automatically converted into a line item. You don't just see a total; you see a granular breakdown of costs avoided through precision adjustments.
🔍
Total Traceability
No "black-box" math. Every dollar saved is linked to a specific batch, a verified mechanism, and a transparent cost-avoidance formula.
The ROI Architecture
By integrating financial modeling directly into the production workflow, Discover provides a dual-layered advantage.
1 Technical Precision
AI builders refine the blend model continuously to minimize waste. Recipe rebalancing happens before the valves open — every variance in the COA receipt translates to a deterministic adjustment, not a corrective top-up after the fact.
2 Financial Validation
The system quantifies saved raw materials, reduced energy, and avoided re-work in real-time. Each adjustment is a Receipt line; the dollar value derives from your plant's actual rates, not from a vendor's brochure.
Lube Oil Blending Plant
Each lubricant in the master carries its own recipe. The Recipe Workbench applies the impact-report financial model against the active recipe and current COA receipt.
Lubricants in master
—
recipe-tagged blends
Annual blend volume
—
—
Total batches / yr
—
across all blends
Expected savings impact
—
USD / year · expected case · traceable in Receipt
Material avoidance rate
$8.25 / MT
avoided correction top-up per blended tonne
COA variance live
—
lubricants with receipt actuals
COA history records
—
total saved receipts across portfolio
The Discover workflow
Five steps. Every step generates evidence; every saving carries its derivation.
Step 1
Configure recipe
Tag each lubricant with components, mass %, and roles (swing / floor / VM / fixed).
Step 2
Capture COA
Receipt actuals per component — KV, additive viscosity, Ca / Zn / P. Predicted blend updates live.
Step 3
Rebalance
Multi-variant solve holds the compliance floor and swings base oils to land target spec.
Step 4
Audit hardware
Predicted vs achieved fingerprint isolates flow-meter calibration vs valve-leak issues.
Step 5
Read the Ledger
Per-batch ledger of avoided corrections — which COA, which mechanism, how many dollars.
Portfolio savings summary — Expected case
Per-lubricant savings derived from the active recipe and any captured COA. Material avoidance + RFT lift; both line items appear in the Savings Ledger with their batch-level derivation.
Lubricant Recipe components Annual MT Material avoidance $ RFT lift $ Total Discover savings COA Status
"
Efficiency is not about working harder, but about removing the barriers to excellence. With Omniblend, the factory floor becomes a symphony of data, where every molecule is optimized, every asset is orchestrated, and every employee is empowered.
Ganesan Ganapathi · Concept Architect · Master Strategy Document, Chapter 5
Lubricant master · recipe configuration
Each lubricant carries its own recipe formula. Click any card to edit components, mass percentages, and roles. The calculator uses these values directly.
Recipe roles drive the math: Compliance-floor components (additive package) cannot be reduced below their licensed minimum. Swing components (base oils) are rebalanced to absorb COA variance. Fixed components hold steady. The financial model uses the additive vs base oil cost differential straight from each recipe.
📥 Bulk upload · Lubricant master & batch ledger
Replace the seeded demo data with your plant's real assessment inputs. The Excel template carries two sheets: Lubricants (recipes, costs, RFT baselines) and Batches (per-batch predicted vs achieved fingerprint, mass, dates). Predictions and savings are recomputed from your data. Download the template first if you haven't built one yet.
What gets imported: uploading replaces the active lubricant master and the captured batch ledger. Plant config and BYOK API keys are not affected. To start with a clean slate, use the Clear data button on the Savings Ledger or click the Reset assessment data… button in the top navigation bar.
Recipe component legend
Swing — rebalanced by Omniblend to absorb COA variance Compliance floor — licensed chemistry, cannot drop below minimum Fixed — held at recipe value Viscosity modifier — adjusted with viscosity targets
🧙 Recipe Workbench · New Batch Wizard
Walk through the five steps to register a new batch. Pick the lubricant, enter batch details, capture the COA (pull from lot inventory or clone a past batch), review Discover's prediction, and commit. The committed batch joins the Savings Ledger with full audit trail.

Step 1 · Pick the lubricant

Select which product is about to be blended. Each card shows recipe component count and recent batch activity. Use the search and category chips to narrow the list.
🔍
Showing all lubricants
Pick a lot
Clone COA from a past batch
Batch drill-down
📒 Savings Ledger
Every batch, every dollar. Each card below is one real batch with a short briefing of what happened and the full computation. The sidebar carries portfolio aggregates for finance reporting.
Every batch in the last 30 days, regardless of outcome.
📊 Generate executive briefing (Plant Leadership Report)
Engine:
⚠ Your API key is stored only in this browser's localStorage.
Ready · click Generate briefing to produce the executive summary.
Briefing not generated yet.
Method note · how the figures here are derived
Material avoidance per batch (only when Discover applied a rebalance) = batch volume (MT) × top-up % × ($/MT additive − $/MT base oil)
RFT lift per batch (when batch lands in spec) = $3.15/MT × actual MT × RFT-gap closure (90% → 99% target)
Captured Discover total = sum of (material avoidance + RFT lift) across all batches in the last 30 days
Annualised projection = captured 30-day total × (365 / 30) — at the captured operating rate

Scenario envelopes (Conservative · Expected · Optimistic) apply 0.67×, 1.00×, 1.33× to the projection.
👁
Focused on
—
All analytics on this page now show this lubricant only.
Expected value pool
—
annual · all lubricants
Confidence range
—
conservative–optimistic
% of material cost
—
expected case
Verdict
—
awaiting computation
📊 Plant Leadership Report · Financial perspective
A narrative briefing for finance and plant leadership: where Discover's savings come from, how they distribute across the portfolio, and which lubricants and scenarios drive the result. Generates from the live simulation. Local generator runs offline; toggle "Use my LLM" to route through Omniblend AI for a richer narrative.
Engine:
⚠ Your API key is stored in this browser's localStorage — only on this device. Calls go directly from your browser to the provider. You'll see usage on your provider account dashboard.
Ready · click Generate report to produce a narrative briefing.
No report generated yet. The finance briefing covers savings totals across all three scenarios, top-contributing lubricants, COA variance status, $/MT realised, and the headroom between conservative and optimistic cases.
Detailed scenario report
Per-lubricant savings across the three scenarios. Click any row to drill into per-batch costs, the recipe-driven loss-category breakdown, and the saved scenario detail.
Lubricant COA Annual MT Material cost ($) Conservative Expected Optimistic Range width % of pool Saving as % cost Drill
Range per lubricant ($/yr) — heat map
Conservative · Expected · Optimistic side-by-side. Cells are rank-colored within each column so skewed distributions don't crush smaller lubricants into invisibility. Hover any cell for the exact figure.
Savings mix — expected case
Material avoidance vs RFT lift across the portfolio. The two components of the Discover total.
Pool composition
Expected-case contribution by lubricant.
Saving rate vs volume
Where size meets opportunity. Top-right lubricants are highest-leverage.
Range-bound cost analysis — Worst / Expected / Best
Annual material cost per lubricant at each scenario. The worst case is the unmitigated baseline; best case is the optimistic-scenario residual cost after Omniblend savings.
Lubricant Annual MT Worst $/MT Expected $/MT Best $/MT Net saving $/MT Saving rate Total cost ($) Annual saving ($)
Portfolio savings heat map
Each cell is colour-coded by savings intensity. Darker = larger annual saving. The slider below is a highlight cap — everything from $0 up to the slider value stays in colour; everything above it dims out. The page loads with the slider mid-track so the filter is immediately visible — click Show all to clear it, or drag to adjust.
Cold
Hot
$0 —
Highlight cap active · drag the slider to adjust, or click Show all to clear
Recipe-driven savings decomposition
How each recipe role contributes to expected-case savings. Swing components carry the rebalance opportunity; compliance floors set the cost differential; fixed components don't move.
Lubricant drill-down
🛠 Generate Operations Briefing
Plant Leadership Report · Quality & Operations perspective. Generates from the live batch data. Local generator runs offline; toggle "Use my LLM" to route through Omniblend AI for a richer narrative.
Engine:
⚠ Your API key is stored only in this browser's localStorage.
Ready · click Generate briefing to produce the operations briefing.
Briefing not generated yet. The operations briefing covers COA variance live across the portfolio, predicted vs achieved fingerprint divergence, RFT gap closure by lubricant, and prioritised actions for the quality team.
🔬 Quality Findings · Per-Batch Fingerprint
Every batch's molecular fingerprint reconciliation in one scan. Predicted-vs-achieved Ca/Zn/P/KV with hardware verdict. Click 🔍 Inspect on any batch to open the deep diagnostic.
How it works: Calcium, zinc and phosphorus are non-volatile — predicted % follows directly from the additive mass-balance. Achieved % comes from the lab. Any divergence isolates either flow-meter calibration or valve integrity.
Every batch in the last 30 days, regardless of fingerprint outcome.
Audit logic reference
Case 1 — Flow meter alert
Lab Calcium > Predicted Calcium (or Zn > predicted) → more additive was physically dispensed than the meter recorded. Schedule additive flow-meter calibration.
Case 2 — Valve leakage
Metals match prediction but viscosity is low → additive dose was correct, so the deficit is excess base oil. Inspect base-oil valves for leakage / bypass.
Quality Findings · batch inspection
Method · how the model works
Three mechanisms drive the 90% → 99% RFT lift, traced from the Strategic Impact Report. Plus the financial formulas that turn each mechanism into a line on the Savings Ledger.
Blind-period elimination
Plants blend on theoretical values and discover variance after the lab test post-blend. Omniblend digitises the COA before the cycle and re-calculates the recipe before execution.
Multi-variant modeling
Manual corrections fix one property and break another. The optimiser solves physical, chemical, and elemental targets simultaneously — first blend hits all specs.
Tribal knowledge → algorithm
Today's 90% RFT depends on a senior blender. The model institutionalises that judgment so 99% is reached regardless of the operator on shift.
Financial model · exact formulas
Per-batch material avoidance (only when Discover applied a rebalance) = batch volume MT × top-up% × ($/MT additive − $/MT base oil)
Per-batch RFT lift (when batch achieves spec) = $3.15/MT × actual MT × RFT gap closure (90% → 99% target)
Captured 30-day total = sum across all batches in the ledger
Annualised projection · expected = captured 30-day total × (365 / 30)
Conservative / Optimistic envelopes = projection × 0.67 / × 1.33
Kettle hours freed = batches × 2 hrs/batch · informational only, not monetised in the Discover total. That belongs in the future Resources product.
Captured savings vs projection · live from the batch ledger
The Discover engine summed across the batch ledger from the last 30 days — and projected forward at the captured operating rate. The reference-plant calibration is shown below to verify the engine reproduces the published savings rates at known operating parameters.
Captured (last 30 days)
—
— batches across — lubricants
Annualised projection · expected
—
Material — + RFT —
Realised $/MT
—
captured rate · vs $1.58/MT reference
Conservative envelope
—
×0.67 around expected projection
Optimistic envelope
—
×1.33 around expected projection
Discover interventions
—
— rebalances + — hardware flags
Reference-plant calibration: at the 25,000 MT reference plant (250 batches/yr at 90% RFT baseline, $2,500 additive vs $850 base oil, 0.5% top-up), the formulas reproduce $39,375 conservative / $78,750 optimistic + 500 hrs freed. The figures above are the same engine run against your captured batch ledger and projected forward at the observed rate.
Constants traced from the impact report
$8.25 / MT material avoidance baseline · $3.15 / MT RFT lift at full gap closure · 90 → 99% RFT target · 2 hrs / batch kettle hours freed (informational)
🧨 Reset assessment data
This toolkit is an assessment workspace. The kill switch below wipes the captured batch ledger, the wizard's in-flight state, and re-seeds fresh demo batches so you can start a clean assessment. Optionally, escalate the wipe to also clear stored API keys, plant configuration, and the lubricant master.
Looking for Excel upload / template? It lives on the Lubricant Master tab — that's where you configure recipes and costs.
For tab-scoped data clearing, use the Clear data button on the Savings Ledger or Quality Audit tabs.
A confirmation dialog will explain exactly what gets wiped before anything happens.
🗑 Clear batch data
This will wipe the captured batch ledger across the toolkit. Every committed batch and seeded demo batch is removed. The Savings Ledger, Quality Audit, and Analytics tabs will reflect the empty state immediately.
The lubricant master, recipes, costs, plant config, and BYOK API keys are not affected. After clearing, you can populate the ledger by committing batches in the Recipe Workbench or by uploading a batch sheet from the Lubricant Master tab.
This action cannot be undone.
🧨 Reset assessment data
This will permanently clear:
  • The captured batch ledger (all live committed batches)
  • The Recipe Workbench's in-flight wizard state
Demo seed batches will be regenerated from scratch.
This action cannot be undone. If you have committed batches you want to preserve, export them first (feature coming soon).
📥 Upload summary