Work

Asked the questions a buyer actually asks, ChatGPT named this site in 57 of 78 checks.

Three properties we build and run, comparing the 90 days to 22 May 2026 with the 90 days to 30 September 2026. Between them: 38.5× search-click growth, 67,687 impressions in 90 days and 967 pages indexed, all from Search Console or a logged AI probe. One of the three has not turned its reach into clicks yet; that is here too.

Read the seven full case studies

These are our own properties, not client work. We build and run all three, and we hold the Search Console access, so every figure can be produced on request.

How do you know these numbers are honest?

Like-for-like 90-day windows

The 90 days to 22 May against the 90 days to 30 September, each from a single Search Console fetch.

Denominators are published

Every AI-citation count comes with how many times we checked. An engine we could not measure is marked unmeasured, not zero.

The unfinished one is on the page

One property has grown its reach without the clicks following. It gets the same space as the two that worked.

No number without a source

Search figures from Search Console; model, test and route counts from the repositories. Nothing estimated.

B2B agricultural exporter, selling internationally

Radiance Overseas

Close to 900 pages generated from structured product data, not hand-built — and 856 of them are indexed.

Clicks from search (90 days): 58 → 3856.6×
Impressions (90 days): 3,213 → 18,7025.8×
Queries ranking: 489 → 1,8523.8×
Pages indexed: — → 856 of 899

Python page generatorsJSON product hierarchiesStatic HTMLHostinger CDNCloudflare DNS

Where AI is: None in the site itself, and it would not help. The pages are generated deterministically from a product catalogue, so they are identical every build and reviewable in a diff. The automation is upstream: keyword research, index coverage, schema validation and content briefs all run on a schedule and write back into this repository.

Read the full case Visit radianceoverseas.com

Student accommodation, competing on local search

Radiance Residency

The steepest curve on this page, from a site that was invisible in May — and ChatGPT now recommends it in 73% of the questions students actually ask.

Clicks from search (90 days): 15 → 57738.5×
Impressions (90 days): 2,849 → 47,49616.7×
Queries ranking: 287 → 1,8516.4×
Named by ChatGPT: — → 57 of 78 checks
Pages indexed: — → 68 of 79

Static HTML, no build stepGA4Google AdsPython bulk-edit scriptsHostinger CDN

Where AI is: A rule-based assistant, and we are calling it that on purpose. The on-site helper matches keywords against a hand-written knowledge base in English and Hinglish — it is not a language model and does not behave like one. An LLM version exists in our own codebase and has not shipped here, so we are not claiming it.

What is not working: Much of the impression growth is informational traffic — queries like “best cafes for studying” — which ranks well and converts poorly. Impressions are not the point; enquiries are, and separating the two is ongoing work.

Read the full case Visit radianceresidency.com

Direct-to-consumer wellness brand

Blossence

Search reach up almost fivefold, and ChatGPT now names it in more than one in five questions about essential oils. Turning that reach into clicks is the live work.

Impressions (90 days): 309 → 1,4894.8×
Queries ranking: 42 → 2205.2×
Named by ChatGPT: — → 18 of 81 checks
Pages indexed: — → 43 of 192

Next.js 16React 19Tailwind 4Headless WooCommerceRazorpayVercel

Where AI is: None in the storefront. The Amazon channel is synced daily through the Selling Partner and Advertising APIs — 51 listings — which is integration work rather than intelligence, and we would rather name it accurately than call it AI.

What is not working: Clicks have not followed the reach yet — 13 in the first window, 14 in the latest, and almost all of them people searching the brand name. The cause is measured rather than guessed: only 43 of 192 known pages are indexed, most of the rest sitting at “discovered, not indexed”, so the new reach is mostly page-five-and-beyond positions that nobody clicks. Indexing is the active workstream.

Read the full case Visit blossence.com

Software

What we build when the answer is not a website.

Three products we designed, built and run. Each says exactly what its AI does, because “AI-powered” on its own has stopped meaning anything.

Screens not shown publicly

Property management for shared housing

WhatsApp-first property management

Running one live property — 31 rooms, 61 beds

Rent, tenancies, KYC, agreements, meter readings and resident messaging, run from WhatsApp rather than an app nobody installs.

70
Data models: 48
API modules: 879
API tests: 32
Migrations

TurborepoNext.js 14NestJSPrismaPostgreSQLReact NativeWhatsApp Cloud API

Where AI is: Vision, doing an unglamorous job well. Residents pay by UPI and send a screenshot; the system reads the amount, reference and timestamp off that image and matches it against the ledger. The same approach reads electricity meter photos, and a matcher reconciles bank statements against expected rent. The work AI removes here is retyping, which is exactly the work that was being done by hand.

What is hard about it: It talks to residents on WhatsApp through Meta's own API rather than a reseller, and it checks the messaging window before sending — so an undeliverable message is reported as undeliverable instead of vanishing. House rules are versioned and hashed, so what a resident agreed to can be proven later.

Read the case study

Screens not shown publicly

Accounting and tax compliance for small businesses

AI bookkeeping and tax compliance

Deployed, in testing — not yet writing to a production ledger

Double-entry books, GST returns and a two-way bridge into Tally, driven by messages a business owner actually sends.

2,429
Backend tests: 224
Backend modules: 63
Migrations: ~105k
Lines of code

FastAPIAsync SQLAlchemyPostgreSQL 17React 18ViteTelegram Bot API

Where AI is: Language in, bookkeeping out. A message like “sold 12 bags to Sharma, 4,800, cash” becomes a correctly classified double-entry voucher, and invoices and receipts are read into structured line items. Every AI output that cannot be justified has to raise a review item rather than guess — an assistant that quietly invents a ledger entry is worse than no assistant.

What is hard about it: The Tally bridge verifies its own work. After writing a voucher it reads Tally's own Day Book back to confirm what landed, rather than trusting the import response. That discipline caught a real bug: GST was being derived from a single hard-coded ledger holding a fraction of the actual tax postings, which understated the returns. Rates are now derived structurally, and the system refuses an off-slab rate instead of rounding it.

No public interface — it runs the pipeline

The search, AI-visibility and enquiry engine behind everything here

Our in-house growth engine

Production — runs the numbers on this page

Every figure in the case studies above is produced by this. It is the system we run client work on, pointed at ourselves first.

195
Database tables: 219
Frozen API routes: 1,824
Enquiries triaged: $2.43
Monthly running cost

PythonFastAPIPostgreSQL with row-level securityCaddyDockersystemd timers

Where AI is: Used where judgement helps and kept away from arithmetic. Models draft outreach, classify incoming enquiries and summarise competitor movement; they never produce a number. The natural-language query interface answers from SQL, so a figure on a dashboard can always be traced to the rows behind it.

What is hard about it: It joins Search Console to Analytics on the landing-page path, which turns “this query got clicks” into “this query produced a session that converted”. That join is the one thing most reporting tools charge separately for, and it is why the numbers on this page are attributable rather than decorative.

Read the case study

Answer engines

Do AI assistants actually recommend these businesses?

Yes, and we can show how often. We ask web-connected assistants real buyer questions and record whether the property is named. Asked what students ask, Perplexity names Radiance Residency in roughly one answer in twelve.

AI citation counts by engine, with the number of probes behind each.
Engine	Property	Cited	Times checked	Rate
ChatGPT	radianceresidency.com27 buyer-intent questions, web search on	57	78	73.1%
ChatGPT	radianceoverseas.com43 buyer-intent questions, web search on	28	147	19.0%
Gemini	radianceoverseas.com28 buyer-intent questions, Google Search on	16	75	21.3%
ChatGPT	blossence.com28 buyer-intent questions, web search on	18	81	22.2%

Last checked 12 October 2026.

Not measured yet: Claude, Perplexity

Only probes run after the 19 September fix are counted here — before it, a failed API call was saved as a confirmed non-citation. Claude and Perplexity have not been re-run on the fixed code, so they are reported as unmeasured rather than as zeros.

Enquiries

1,824 enquiries captured and triaged.

Since 16 April 2026, from email and WhatsApp, deduplicated and moved through one pipeline. A capture-and-triage number, not a sales figure.

1,439

385

1,264

New

482

Contacted

Negotiating

Quoted

Tell us what you need. We’ll reply with a price.

One business day, one straight answer. No discovery call unless you want one — and if we aren’t the right fit, we’ll say so.

Get a quote See pricing