One view of every store's power bill for retail and restaurant chains
TL;DR Ampshelf collects the electricity bills of every outlet in a retail, pharmacy or restaurant chain, whatever the DISCOM, and ranks the stores against each other. It flags wrong tariff categories, oversized sanctioned loads and stores whose ACs keep running after closing. The chain's facilities head pays a monthly fee per store.
Commercial users drew about 135 TWh of India's 1,623 TWh of electricity in FY 2024-25, roughly one unit in twelve[1]. A good chunk of that sits in chain outlets, and one grocery chain alone passed 500 stores in April 2026[2].
Each outlet is a separate connection with its own DISCOM, bill format, due date and tariff category. In most chains an accounts clerk pays these bills and nobody compares them.
So a store paying for twice the load it uses, or one leaving signage and ACs on overnight, looks exactly like every other line in the ledger. Nobody notices.
Ampshelf starts with the bill, not the building. Stores upload a photo or PDF of each bill, or Ampshelf pulls it from DISCOM portals with the chain's permission, and a parser turns dozens of bill formats into one table. Then it does three things.
It ranks like-for-like stores by units per square foot per trading hour, so the worst 10% stand out. It checks each bill for fixable money - sanctioned load far above recorded demand, a commercial store billed in the wrong category, late-payment charges and arrears that never cleared. And for the worst stores only, it fits a clip-on meter for two weeks to show when power gets used, then hands the store manager a short list of what to switch off and when.
Ampshelf doesn't control equipment. If a chain wants remote AC control then it can add a vendor for that.
The seed pitched a broad AI platform. We kept the goal and narrowed it to chains, because their pain is lots of small, scattered bills, and large single buildings already have automation vendors.
The user is a regional facilities or store-operations manager. The payer is the head of facilities or the CFO, who signs off the power bill line.
Pricing would be per store per month, plus a share of any refunds won on billing errors. We found no published price list for Indian multi-site energy software, so test prices in the sprint.
Chains are adding stores fast[2], and every new store is one more small bill. India's distribution reform scheme is pushing smart meters, which will make interval data easier to get[1]. The DPDP Act passed in 2023 sets clear rules for handling account data[3]. And cheap clip-on meters mean you no longer need to wire every store to learn something useful.
We found no published count of how many electricity connections Indian chains hold, and no data on how often their bills are wrong. The bottom-up view - commercial demand of about 135 TWh a year[1], and single chains running hundreds of outlets[2].
Treat the commercial total as adjacent context, not as your market.
Building-automation firms owned by global electrical majors and Indian energy-as-a-service players[4] sell building automation and energy savings to commercial sites, mostly with hardware in every building. Meter-analytics vendors sell to utilities, with a small-business line, but not to chains[5].
The gap is a light, bill-first tool that a chain can roll out to 300 stores in a month without wiring every one.
| Difficulty | Moderate (36/100) |
| Starting capital | ₹60–110 L |
| Time to launch | 4–6 months |
The software is ordinary. The grind is bill parsing across dozens of DISCOM formats and getting store managers to change their habits. Refunds for billing errors need chasing with each DISCOM, which is slow work. Expect chains to run a paid pilot of three to six months before they commit.
No sector licence. Analysing a customer's own bills needs no permission from the electricity regulator.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Don't let AI file a refund claim or change a tariff category on its own. Every money decision needs a person to check the bill first.
Want to work out the AI layer for Ampshelf? Talk to Yash →
Here's what can go south -
Good for a founder who has run facilities or finance at a chain and can open doors to three or four of them. If you're hoping AI alone will find the savings, I'd pass.
First prove one thing - that bill checks alone find recoverable money in at least one store in five.
Plan for roughly ₹60–110 L, mostly on bill parsing software, a small field team and paid pilots.
No. You need standard company and GST registration, and a data processing contract with each chain under the DPDP Act.
Yes, chains with scattered outlets exist everywhere. Check how local utilities share bill data and what your data protection law requires.
The free Ampshelf validation kit has the line-by-line capital breakdown, the effort scoring, the data and billing rules checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for facilities heads, and 5 assumptions to test. Download the kit →
Thinking about building Ampshelf or something close to it? The make-or-break work is showing a chain real money in its own bills within the first month, and I'd be glad to help you plan that pitch. If you need a hand, you know where to find me. Talk to Yash →
For illustration only. Names are fictional and ideas hypothetical; sourced facts are real but may be out of date. Nothing here is legal, financial or professional advice, and Yash Khatri is not personally liable for any use of it. Read the full disclaimer.