Synthetic case explorer

Lattice Forecourt

Make a price proposal explainable

Follow three local specialist checks into a deterministic pricing gate.

Explore the evidence

Select a precomputed synthetic case, inspect input assumptions and download its JSON evidence or result CSV.

Open Lattice Forecourt explorer →

What this edition demonstrates

Integer mills, market median, margin floor, change cap and rounding tick.

No prices are applied. These are local rules, not model-agent calls.

The public website hosts a read-only synthetic projection. The full authenticated source application and its Git history are maintained privately.

All fixtures are fictional. No sign-in, customer records, operational writes or paid external model calls are involved.

Publication notices · Application metadata

Explore the application directory

MATHEMATICAL SPECIFICATION & UNITS

Governing Equations, Inputs, Outputs & Dimensional Conventions

Lattice Forecourt executes deterministic client-side calculations in the browser. Every scenario is governed by explicit mathematical formulations and unit checks:

p_market = median(M); p_floor = c_mills + m_min; p_capped = clamp(p_target, p_0 − Δ_max, p_0 + Δ_max); p_final = round_to_tick(max(p_capped, p_floor), τ_mills)

Declared Units & Arithmetic Conventions: All monetary prices, costs, margins, caps, and rounding ticks in integer mills per litre (1,000 mills = 1.00 currency unit); volume in millilitres (mL).

Scope & Operational Boundary: Deterministic fuel-pricing workflow with specialist checks, margin floors, change caps and rounding. No live AI/model agents or actual commercial price changes.

Launch Interactive Lattice Forecourt Runtime → · Read Full Methodology & Validation Notes → · Download JSON Dataset (run/catalog.json)

REPRODUCIBLE WORKED EXAMPLE & BENCHMARK TABLE

Verified Case Study & Held-Out Reference Table

Worked Verification Case: Case 1 (Current price p_0 = 1,600 mills/L, wholesale cost c = 1,400 mills/L, competitor market = [1,610, 1,640, 1,650] mills/L, min margin = 100 mills/L, max change cap = 100 mills/L, rounding tick = 10 mills/L): Market median is 1,640 mills/L. Margin floor requires ≥ 1,500 mills/L (passed). Change cap allows [1,500, 1,700] mills/L (passed). Tick rounding at 10 mills preserves 1,640 mills/L, so the deterministic gate status is 'Clear' with Proposal = 1,640 mills/L.

Reproducible benchmark scenarios shipped in the Lattice Forecourt browser edition (Synthetic / Precomputed Evidence)
Case ID	Cost (mills)	Current (mills)	Market Median	Floor / Cap Rule	Proposed (mills)	Gate Status
Case 1 (Standard Clear)	1,400	1,600	1,640	Floor 1,500 / Cap ±100	1,640	Clear
Case 2 (Cap Bound)	1,450	1,580	1,720	Floor 1,550 / Cap ±80	1,660	Capped
Case 3 (Floor Conflict)	1,560	1,600	1,620	Floor 1,680 / Cap ±50	Blocked	Floor vs Cap Conflict

PROCEDURAL USAGE GUIDE

How to Use Lattice Forecourt in 3 Reproducible Steps

Select a Baseline Scenario or Practice Module: Open the interactive browser runtime and choose one of the shipped synthetic presets or diagnostic modules to load its verified input parameters and unit declarations.
Inspect Intermediate Traces & Constraints: Review the computed ledger, error diagnostics (MAE/RMSE/margin/balance checks), or bilingual explanations alongside the governing equations documented on khanbot.co.technology/methodology/.
Export Portable Evidence (JSON / CSV): Export your scenario results, assumptions, or local study plan as a self-contained JSON or CSV file for offline verification and local re-import.

FREQUENTLY ASKED QUESTIONS

Methodological & Boundary FAQs

What mathematical model and units does Lattice Forecourt use?

Lattice Forecourt applies p_market = median(M); p_floor = c_mills + m_min; p_capped = clamp(p_target, p_0 − Δ_max, p_0 + Δ_max); p_final = round_to_tick(max(p_capped, p_floor), τ_mills). Units: All monetary prices, costs, margins, caps, and rounding ticks in integer mills per litre (1,000 mills = 1.00 currency unit); volume in millilitres (mL).

What are the operational boundaries and limitations of Lattice Forecourt?

Deterministic fuel-pricing workflow with specialist checks, margin floors, change caps and rounding. No live AI/model agents or actual commercial price changes. All computations run deterministically in the browser on synthetic or precomputed demonstration fixtures.

How can I reproduce the benchmark results in Lattice Forecourt?

Open the interactive browser workspace at ./run/, select Case 1 to inspect the worked result (Case 1 (Current price p_0 = 1,600 mills/L, wholesale cost c = 1,400 mills/L, competitor market = [1,610, 1,640, 1,650] mills/L, min margin = 100 mills/L, max change cap = 100 mills/L, rounding tick = 10 mills/L): Market median is 1,640 mills/L. Margin floor requires ≥ 1,500 mills/L (passed). Change cap allows [1,500, 1,700] mills/L (passed). Tick rounding at 10 mills preserves 1,640 mills/L, so the deterministic gate status is 'Clear' with Proposal = 1,640 mills/L.), and export the JSON or CSV evidence ledger.

PROVENANCE, ATTRIBUTION & RESEARCH ECOSYSTEM

Engineering Attribution & Interconnected Research Hub

Authored and maintained by Mohammad Rezwan Khan, PhD (Aalborg University · BESS QA/QC & Power Systems Specialist). Canonical URL: https://khanbot.co.technology/ · Source revision: verified-rel · Runtime SHA-256 prefix: verified-sha256 · Repository provenance: https://github.com/mohammadrezwankhan/lattice-forecourt (private source engine; public deterministic browser projection).

Application Metadata (app.json) · AI / LLM Discovery Guide (llms.txt) · Full Technical Corpus (llms-full.txt) · Semantic Knowledge Graph (knowledge-graph.jsonld)