
Building the operating system for fresh food commerce.
An artificial intelligence and machine learning platform that makes fresh food operations more predictable, profitable, and less wasteful — grounded in the real-world operating experience of Royal BP Corporation's 43 stores and restaurants.
Royal Fresh is a co-owned product of Royal BP Corporation and Hardik Choksi, who licenses the underlying technology to the group. Post-pilot, the platform will be further developed and supported by Blue River Labs Inc. — the startup he founded.
Across 43 stores and restaurants, the same problem repeats daily: operators make high-stakes decisions about what to buy, prepare, move, mark down, and sell — without knowing the true state of inventory, the real shape of demand, or how much shelf life remains.
This is not a data problem. It is a control problem. Traditional enterprise software was never built for it.
Recorded ledger balances almost always differ from sellable shelf inventory. The system believes stock exists when the shelf is empty.
POS records show only what was bought — missing latent demand when the shelf was empty. Models trained on censored sales learn that demand dropped to zero.
A tray of prepared food exists physically, but its economic utility decays by the hour. Time converts uncertainty into loss.
We do not just aggregate silent data dumps. Royal Fresh accumulates labeled decisions, human overrides, local operating context, and observed physical consequences — each accepted recommendation, count, stockout, waste event, and sale compounds into a better understanding of fresh food.
Independent 2026 food-system research estimates nearly one third of the United States food supply is wasted — valued at roughly $380 billion. Retail alone generated approximately 3.98 million tons of unsold food valued at $26.9 billion.
Waste is born long before disposal — when demand is misunderstood, inventory overstated, or production scheduled against distorted signals. Disposal is the final symptom of an earlier decision.


Perishable departments with variable decay, invisible stockouts, and volatile local demand.

Thin margins, micro spaces, and rapid throughput — where a single spoilage event erases a day's profit.

Guest demand translated into prep tasks, batch timing, and yield-aware ingredient consumption.
Royal Fresh was conceived, designed, and built by Hardik Choksi.
It took shape in early 2026 as an intelligence layer for high-decay commerce — written by its inventor, then brought to Royal BP Corporation, where 43 live stores and restaurants became its proving ground under the leadership of Tushar Patel, CEO.
Conceived the platform, designed the closed-loop architecture, and developed the models. He retains ownership of the underlying intellectual property.
Royal Fresh is a co-owned product of Royal BP Corporation and Hardik Choksi. The corporation operates the brand under a technology license granted by Mr. Choksi.
After the initial pilot phase, Royal Fresh will be further developed and supported by Blue River Labs Inc. — the technology startup founded by Hardik Choksi — through nationwide rollout across grocery and convenience operators, and out to the farmers and partners across the supply chain.
Royal BP Corporation supports the pilot phase within its own operations. Everything after it — the all-out rollout of Royal Fresh across grocery stores, convenience stores, and the wider fresh supply chain — will be engineered and supported by Blue River Labs Inc. (blueriverlabs.ai), founded by Hardik Choksi.
The complete Royal Fresh research and technical briefing document — all 35 pages, as a flip-through book. Click either page (or use the arrow keys) to turn.
Research & Technical Briefing · 2026
Prefer a copy? Use “Open in new tab” or download the PDF above.