--- title: "TenantBill" headline: "TenantBill — Ali Ahmed, Head of Product & Platforms, BijliBachao.pk" description: "Ali Ahmed built TenantBill: photo-verified tenant billing for Pakistani multi-tenant properties — a five-link chain of custody that makes a meter reading defensible enough to bill." canonical_url: https://alioahmed.com/work/tenantbill type: case-study role: "Head of Product & Platforms, BijliBachao.pk" date_label: "2026 – present" status: live date_modified: 2026-08-25 author: name: "Ali Ahmed" url: https://alioahmed.com/ orcid: https://orcid.org/0009-0007-4265-3295 key_figures: - "5 consecutive months closed" - "668 readings captured" - "100% of them carry a photograph" - "0.45% needed a retake" entities: - name: "TenantBill" type: SoftwareApplication - name: "BijliBachao.pk" type: Organization url: https://bijlibachao.pk/ - name: "TenantBill chain of custody" type: DefinedTermSet - name: "Water and Power Development Authority" type: GovernmentOrganization url: http://www.wapda.gov.pk/ same_as: - https://www.wikidata.org/wiki/Q7973437 - name: "Lahore Electric Supply Company" type: Organization url: https://en.wikipedia.org/wiki/Lahore_Electric_Supply_Company same_as: - https://www.wikidata.org/wiki/Q15982516 - https://en.wikipedia.org/wiki/Lahore_Electric_Supply_Company - name: "Google Cloud Vision AI" type: SoftwareApplication url: https://cloud.google.com/vision/ same_as: - https://www.wikidata.org/wiki/Q59898818 person_same_as: - https://www.linkedin.com/in/alioahmed/ - https://github.com/alioahmed - https://orcid.org/0009-0007-4265-3295 - https://x.com/Alioahmed_ - https://about.me/alioahmed - https://dev.to/alioahmed - https://medium.com/@alioahmed - https://hashnode.com/@alioahmed - https://substack.com/@alioahmed - https://huggingface.co/alioahmed - https://bsky.app/profile/alioahmed.bsky.social - https://mastodon.social/@alioahmed - https://www.kaggle.com/alioahmed - https://www.producthunt.com/@alioahmed - https://www.patreon.com/cw/alioahmed - https://www.goodreads.com/user/show/202106301-ali-ahmed - https://fueler.io/alioahmed - https://gitlab.com/alioahmed - https://linktr.ee/alioahmed - https://topmate.io/alioahmed related: - https://alioahmed.com/work/wattey - https://alioahmed.com/work/sts - https://alioahmed.com/work/bijli-bachao full_corpus: https://alioahmed.com/llms-full.txt site_index: https://alioahmed.com/llms.txt --- # TenantBill > TenantBill is a photo-verified meter reading and tenant billing platform built by Ali Ahmed, in the Wattey family at BijliBachao.pk. A mobile app captures every reading with a timestamped photograph of the meter face and reads the number from it by OCR; a portal validates each reading against its own image, generates every bill from one published formula, tracks arrears, reconciles the tenant bills against the single utility bill, and closes the month into a frozen snapshot. It has closed five consecutive monthly cycles at a shopping mall in Punjab, across 668 readings of which every single one carries a photograph. Its product insight is that the deliverable is not the invoice, but the evidence that makes the invoice unarguable. **Canonical URL:** https://alioahmed.com/work/tenantbill **Role:** Head of Product & Platforms, BijliBachao.pk — 2026 – present **Summary:** Ali Ahmed built TenantBill: photo-verified tenant billing for Pakistani multi-tenant properties — a five-link chain of custody that makes a meter reading defensible enough to bill. ## Key figures - **5** — consecutive months closed - **668** — readings captured - **100%** — of them carry a photograph - **0.45%** — needed a retake ### A bill is only worth what it can survive. A Pakistani shopping mall has one utility connection and hundreds of units behind it. The flagship is a roughly 545-unit property in Punjab, with 129 shops onboarded into the system and about 111 meters billed every month. Somebody walks the building with a notebook writing down meter numbers; somebody else retypes them into a spreadsheet, where a formula nobody can audit turns them into bills. Once a month a shopkeeper says the reading is wrong, and there is no proof either way. The tenant bills almost reconcile against the utility bill — almost — and the gap is the owner’s loss. This is the quiet leakage in every multi-tenant property: disputes that cannot be settled, a shortfall that cannot be explained, and hours of spreadsheet work producing a number nobody fully trusts. TenantBill’s answer is not a better invoice. Any spreadsheet can produce an invoice. What a building manager actually lacks is the ability to prove one — so the system was built backwards from the dispute, and the invoice falls out of the evidence rather than the other way round. ### A photograph is not proof. A chain is. Attaching a picture to a meter reading is the easy idea, and it is where most photo-verified billing stops. It does not survive a real dispute, because a photograph taken and approved by the same person proves nothing a handwritten number did not already prove. What holds is a chain of five links, and removing any one of them collapses the other four back into a spreadsheet. Every reading is captured with a timestamped image of the meter face, retained permanently. Optical character recognition lifts the number off that image and records how confident it is, so a poor read is flagged for review rather than quietly trusted. The field operator who captures a reading cannot approve it — the app permits capture and nothing else, no rates, no billing, no approval. A wrong reading is corrected by an audited retake rather than an overwrite, so the original stays in the record and the correction is visible as a correction. And at month end the figures freeze. That interdependence is the engineering insight. A photo alone is theatre if the same hand can bless it. Separation of duties is theatre if a number can be silently overwritten. An audit trail is theatre if the closed month can be quietly amended. ### The month has to close, and four numbers have to tie. “The numbers never quite add up” is not a complaint about arithmetic. It is that the tenant invoices and the utility bill describe two different worlds, and the gap between them is the owner’s loss with no name on it. A multi-tenant property has four quantities that must reconcile every month: what was bought from the utility, what was generated on the roof, what was billed to tenants, and what the common areas consumed. TenantBill ties them in one ledger — solar credited against building consumption without being counted twice, and shared lighting and lifts billed to the building rather than back to a shop. That reconciliation is what turns billing software into a book of account. Then it stops moving. At close the readings, bills and payments become read-only; reopening is possible, because sometimes it must be, and is recorded against a named role and a timestamp. A set of books that can be quietly amended after everyone agreed to it is not a record. The states a real property is actually in are on the page too, because every naive schema breaks on them: a shop between leases that still has a meter and still accrues common-area load, tenants registered without a unit who must not receive a bill, and shops deactivated without being deleted so their history survives. ### What the production record actually shows. Read-only queries against the production database on 25 August 2026 settle what this chapter can claim, and the measured version is stronger than the marketing one it replaces. Five consecutive monthly cycles have been closed — March through July 2026 — each in the first week of the month following, with a sixth in flight. That regular cadence is the difference between a pilot and a production system. The first reading was taken on 5 March 2026 and the platform has run continuously since. Across those cycles 668 readings have been captured, and 668 of them carry a photograph. Not almost all — all of them. The figure also checks itself: 668 is roughly 111 active meters times six cycles, which is exactly one reading per meter per month, so the volume, the meter count and the cycle count each corroborate the other two. The retake workflow is demonstrably used and rarely needed: three retake requests against 668 readings, about 0.45 per cent — one fulfilled, two cancelled after review. The cancellations matter as much as the fulfilment, because they show triage happening rather than a workflow nobody touches. Two months have genuinely been reopened after closing, in March and April 2026, and the ledger retained the complete record through both — the claim that reopening is survivable is not a design intention but something the system has actually done. Sixty-one closed bills have since been settled as arrears, so the carry-forward loop is exercised rather than theoretical. Seventeen of the ninety-eight meter-holding tenants hold more than one meter, and the largest holds five — which is why consolidated invoicing exists at all. ### Where the chain ends: the evidence prints on the bill. Every guarantee described above happens inside the system. The bill is where the evidence leaves it and lands in the hand of the person being charged — which is why the platform team, asked what the single strongest thing about the product is, names this document first. Two photographs of the meter print on the bill itself: a cropped view of the display and the full meter face, both timestamped. A tenant disputing the figure is not asked to trust a database; they are holding the picture. The correspondence is checkable on the paper — the display in the photograph reads 24429, and the bill states a current reading of 24,429. The step most likely to be disputed is the one most visibly worked out. A CT-metered connection reads a fraction of the real current and has to be multiplied up, so the bill prints the arithmetic in full: a meter difference of 445 multiplied by a factor of ten gives 4,450 effective units. Arrears arrive the same way — two carried months and a late fee itemised above the current charge, with the grand total showing its own components, and the amount written out in words in the Lakh convention beneath. A tear-off payment slip completes it. The bill shown on this page is a demonstration produced by the production generator: the template, styles, arithmetic and amount-in-words are the product’s own output, taken verbatim from its billing code, and only the input data differs. The tenant, shop and property are fictitious and belong to nobody. The meter photographs are real. ### The question a spreadsheet cannot ask. Every other guarantee in this system makes a bill defensible. One of them makes the whole building legible, and it is the piece an outsider never guesses. Line loss is what the utility supplied minus what every shop meter recorded — the energy that entered the property and was billed to nobody. Theft, a failing meter, a load somebody never metered. On a spreadsheet it cannot be computed at all, so the owner absorbs it every month without knowing it is there. The power ledger sets the whole property out as one equation: what came in, from the utility and from the roof; what the shops used; what the owner’s own common areas used, charged to the owner and never to tenants; what leaked, colour-graded; what was billed against what was actually collected; and finally net recovery against the utility bill — the building’s real profit. The platform team names it, unprompted, as the strongest screen the product has, alongside a bill that carries the meter photograph on it. ### Four decisions, each removing a class of error. The billing page is read-only by design. Bills derive from readings and rates, so there is no field to edit and nothing to fudge — a whole category of dispute, and of fraud, removed by declining to offer the control in the first place. Rates resolve per shop, then per tenant, then to a property default, so an anchor tenant on a negotiated rate and an ordinary shop are both billed correctly from one system rather than from a parallel spreadsheet of exceptions. Heavy three-phase connections meter only a fraction of the real current and have to be multiplied up by a CT factor — and that multiplication is printed on the tenant’s own bill, so it can be checked rather than taken on trust. Deactivating a tenant carries their shops and meters with it, and nothing is ever destroyed, so a property that changes shape does not lose its history. ### One formula, published — and four roles that separate. Consumption is the current reading minus the previous one. The amount is consumption times the per-unit rate times the meter’s multiplying factor. Tax applies to the amount, and the total adds arrears and any late fee. That is the whole calculation, it is identical for every shop, and the tenant can see it. Transparent arithmetic is what ends a dispute; a total on its own restarts one. Transformer-stepped meters carry a multiplying factor applied in the bill line rather than in somebody’s head, and a tenant holding several units receives one consolidated invoice with a line per meter. Four roles exist and the gaps between them are the audit story: a super administrator who can close and reopen months, an administrator who manages billing and approvals, a billing manager who generates bills and closes the month, and an operator who can capture a reading and do nothing else. The capture app works offline and synchronises when signal returns, with duplicate prevention holding inside the offline queue, because basements and service corridors are the normal case in this building type rather than the exception. Three weeks takes a property from signed contract to a live cycle. ### Where this sits — a small, fast category inside a very large one. The exact category TenantBill occupies, tenant billing software, was worth roughly USD 455 to 460 million in 2025 and is growing 8 to 9 per cent a year — two research firms, Mordor Intelligence and 360iResearch, landing independently within about one per cent of each other. A third house quotes a materially larger figure on a wider definition; this page ignores it, because reaching for the biggest available number is how a claim gets taken apart by the first person who checks it. The platform team puts the same point in one sentence: sub-metering hardware exists and utility software exists, but nothing served the owner in the middle who buys bulk and re-bills retail. That seat is the product’s whole position. That category sits inside a utility sub-metering solutions layer of around USD 3.8 billion, itself inside an electric sub-meter hardware market an order of magnitude larger again. The asymmetry is the argument: meters are a large and crowded hardware market, and the software that turns their readings into bills nobody disputes is the small, fast-growing, genuinely difficult slice of it. Asia-Pacific is the fastest-growing region for the category, at about 7.8 per cent a year through 2032, and the driver analysts name for its commercial segment is tenant billing itself — with commercial buildings holding 28.4 per cent of the market on exactly that basis. Ali did not build this adjacent to where the category is growing; he built it in the fastest-growing region for it, against the named driver, in one of the hardest energy environments in the world. Industry estimates put the leakage in manually billed buildings at 3 to 8 per cent of total energy cost, through reading gaps, calculation errors and tenants underpaying without challenge. That figure comes from a vendor rather than independent research and is directional only. What makes it worth citing is that the causes listed beside it — human error in manual reading, spreadsheets missing edge cases, no timestamped record to show a tenant who disputes a bill, and common-area energy as the biggest source of argument — read as a specification for the product described on this page. ### What the billing layer unlocked. BijliBachao sells a service it calls Solar Share System, described in its own catalogue as a first in Pakistan: commercial solar EPC bundled with TenantBill, where the building owner installs the array and every tenant receives a photo-verified bill crediting solar units against grid units. Read the dependency in that description and the argument is finished. A landlord cannot sell rooftop generation to their own tenants on the strength of a spreadsheet — the tenants have to accept the bill, or the arrangement is a monthly argument about money. The billing layer is not a supporting feature of that service; it is the precondition for it existing at all. Without a bill the tenants accept, the roof stays a roof. That is the checkable version of what changed because Ali was there: he built the thing that made a commercial service line possible. ### What transfers, once you take the meters out. Strip the electricity out and TenantBill is a machine for capturing evidence in the field, reading it automatically without over-trusting the read, and turning it into a figure two parties who disagree can both rely on. Six things generalise. Capture evidence rather than data, so the artefact that proves a value is recorded at the same instant as the value. Make the automation declare its own confidence, because an extraction that cannot say how sure it is will be trusted uniformly, including when it is wrong. Never let one hand both capture and approve. Correct by appending rather than overwriting, so a mistake stays visible. Close the period and make reopening an attributed act. And assume no signal, because field work happens where there is none. That is stock audit, clinical data capture, insurance inspection and proof of delivery. The domain is incidental; the chain of custody is the product. ### What this page does not claim. It runs on one property. Everything here describes a single deployment, and no portfolio is implied. The screenshots are redacted. Every raw capture contained real tenant personal names, shop trading names, meter serial numbers and the operator’s email address; one showed the property owner’s full monthly profit and loss. The tenants are small shopkeepers who never agreed to appear on a portfolio, so their names, their meters and the property’s financial totals are covered. The property may be named — Mall of Muridke — because the platform team names it in material prepared for publication; the tenants inside it may not, and that line is absolute. No rupee figure from this deployment is published anywhere, and no tariff rate appears. The scale figures are an April 2026 snapshot, corroborated by the product’s own screens — the dashboard rendered them on the same date. They disagree with the summary currently published on this site’s Bijli Bachao chapter, and that discrepancy is recorded for a database answer rather than resolved by choosing the more flattering figure. The four outcome percentages sometimes quoted for this deployment — fewer billing errors, faster invoicing, fewer complaints, less administrative work — are the client’s own reported figures rather than instrument readings, and whether any of them were actually measured is an open question. They are therefore not stated as results. ## FAQ ### What did Ali Ahmed actually build in TenantBill? The whole system: the mobile capture application, the OCR pipeline that reads a meter face and scores its own confidence, the approval workflow that separates capture from validation, the billing engine and its single published formula, the arrears and payment tracking, the reconciliation against the utility bill, and the month-close that freezes the books. He leads it as Head of Product & Platforms at BijliBachao.pk. ### What was the real insight behind this product? That the deliverable is not the invoice. Any spreadsheet can produce an invoice; what a building manager actually lacks is the ability to prove one. So the system was built around evidence — a photograph retained behind every reading, a formula shown to the person being billed, and a month that can be frozen — and the invoice falls out of it. Getting that ordering right is the product decision; everything downstream is ordinary engineering. ### Why is capture separated from approval? Because a reading that the same person can both take and approve is not evidence of anything. An operator in TenantBill can capture and nothing else — no rates, no billing, no approval — and a manager reviews each reading against its own photograph and last month’s figure, requesting an audited retake rather than editing a number silently. Without that separation the photographs would be decoration. ### Why are the screenshots on this page redacted? Because the raw captures contained real tenant names, shop trading names, meter serial numbers, the operator’s email address and, on one screen, the property owner’s monthly profit and loss. The tenants are small shopkeepers who never agreed to appear on a portfolio. The screens are shown because they are real and running; the people in them are not the evidence, and neither is the client’s margin. ### Which figures on this page are uncertain? Very few, and that is recent. Read-only queries against the production database on 25 August 2026 replaced the page’s earlier figures, several of which were wrong — the scale twice over, and the optical-character-recognition accuracy twice over. Four outcome percentages that had circulated in marketing material were traced to a document listing them as typical results for a hypothetical property, with no baseline and no measurement anywhere, and were dropped rather than softened. Two billing-cycle claims failed the same check and were dropped too. No speed figure is stated for the automatic reading because no timing telemetry is stored, and no offline-capture count is given because synced readings cannot be distinguished after the fact. What remains is measured. ### What did this software actually make possible commercially? A rooftop with a customer. BijliBachao sells a service it calls Solar Share System — described in its own catalogue as a first in Pakistan — which is commercial solar EPC bundled with TenantBill: the owner installs the array and each tenant gets a photo-verified bill crediting solar units against grid units. A landlord cannot sell rooftop generation to their own tenants on the strength of a spreadsheet, because the tenants have to accept the bill. The billing layer is the precondition for that service existing, not a feature of it. ### How big is the market for a system like this? The exact category — tenant billing software — was worth roughly USD 455 to 460 million in 2025 and is growing 8 to 9 per cent a year, with Mordor Intelligence and 360iResearch landing independently within about one per cent of each other. It sits inside a sub-metering solutions layer of around USD 3.8 billion and an electric sub-meter hardware market an order of magnitude larger again. The meters are a big market; the software that makes their readings defensible is the small, fast, difficult part. Asia-Pacific is the fastest-growing region for the category at about 7.8 per cent a year, and the driver analysts name for its commercial segment is tenant billing itself. Every figure, its confidence grade, and what was searched for and not found is recorded in the repository at docs/research/tenantbill-market.md. ### What does manual tenant billing actually cost a building? Industry estimates put revenue leakage in manually billed buildings at 3 to 8 per cent of total energy cost — consumption that happens but is never billed, through meter-reading gaps, calculation errors, and tenants underpaying without being challenged. That figure comes from a vendor of billing software rather than from independent research, so it is directional and this page attributes it wherever it appears. What makes it worth citing is that the causes named alongside it read as a specification for what TenantBill does: human error in manual reading, spreadsheets missing edge cases, no timestamped record to show a tenant who disputes a bill, and common-area energy as the single biggest source of argument. ### What transfers from this to a problem that is not about electricity? The shape is general: field staff capturing data on a phone in places with no signal, an automated read of that capture that has to declare its own confidence, an approval path where the capturer is not the approver, a computation that has to be explainable to the person it charges, and a periodic close that makes the record immutable. That is stock audit, clinical data capture, insurance inspection and field logistics. The electricity is incidental. --- Rendered from the structured content at https://alioahmed.com/work/tenantbill. Facts trace to the canonical bio; render, don't reword.