The work

Buying gold as an investment in Iran means trusting a shop, a price you cannot verify, and a buyback promise nobody wrote down. The trust gap — not the technology — is the market.

Architected a full investment platform: buy, sell, hold, buyback, pricing, invoicing, delivery and account journeys — with the buyback obligation designed in from day one, not bolted on.

Never launched. The parent group's priorities moved to physical operations before the platform shipped. The architecture and the market model survived and were folded into Shalize's operating system.

A physical gold and silver business running on WhatsApp messages, memory and trust. Volume was climbing faster than the process could hold it.

Rebuilt the operating system end to end — enquiry, order, invoicing, payment verification, fulfilment, follow-up and buyback — and put a CRM under it.

Daily volume moved from 0.5 → 10 kg/day. 4,000+ verified purchasing customers.

Jewellery is not bought on specification. It is bought on meaning — and the category was selling grams.

Positioned Goldoon as story-first: collections, gifting, packaging, charms and personalisation designed as narrative objects, not SKUs.

Shipped as the group's brand-led retail line, with a distinct customer journey from the bullion side of the house.

A business tends to act on the symptom it noticed first rather than on what is actually wrong.

Founded it as the first product out of Whalya, and I own the product: a self-assessment across eight domains, deterministic scoring, an AI-written analysis checked by an automatic reviewer, and a report with a health stage and a 30-, 60- and 90-day action plan.

Live in a free launch phase. It is a self-assessment, not an audit or a certificate. whalyon.com

Small businesses decide what to publish from unsourced AI output rather than from evidence about themselves and their competitors.

Co-founded it; I run product and take part in building it. It collects evidence about the business and its competitors, and a deterministic gate stops any claim whose source the business does not hold, before a person decides.

Deployed; a small pilot with a handful of businesses comes before any sale. vohu.app

An IELTS candidate outside a major city has no examiner, no feedback, and no way to know why their writing scores what it scores.

Built an AI feedback engine that simulates the exam, analyses the response and returns specific, actionable, examiner-shaped feedback — then designed the improvement loop around it.

MVP validated with real submissions and first paying customers — the first proof that people would pay for machine feedback. Wound down in 2025.

Teenagers do not stop reading because books are bad. They stop because reading is solitary, unrewarded and invisible to their friends.

Built a social, gamified reading platform — story journeys, points, medals, narrator characters, quizzes — and launched HisTory as a product-led growth campaign that made registration, profile completion and e-book purchase one continuous motion.

Launched in roughly three months. Partnered with the Ministry of Education and the national Shad platform.

Charitable giving fails on both ends: donors cannot verify need, and social workers cannot prove outcome. Trust collapses in the middle.

Designed the Mehrabani Card — one object carrying the verified need, the beneficiary's story, the donation and the outcome — and built Soha underneath it as the verified-needs data layer.

Built for national-scale distribution. The impact-reporting loop closed the donor–beneficiary trust cycle.

A video-social product for an audience that had no safe, culturally native place to make and watch short video.

Led it from design through launch — content consumption, engagement, retention and the creator workflow.

300K users in the first two months.

Parents drown in parenting advice and starve for parenting guidance. The gap is between expert knowledge and a decision at 11pm on a Tuesday.

Designed taxonomies, need classifications and content pathways that turn specialist knowledge into something a tired parent can act on.

Shipped and iterated over four years as Aghigh's most human-centred product.

Adolescents are treated as an audience to be broadcast at, not as participants with agency.

Defined the product problem and designed engagement flows that turned teenagers from consumers into contributors.

Launched and validated the early model. I left before it scaled — I am only claiming the zero-to-launch work, not its long-term trajectory.

The Ghadir Highway corridor existed only as corridor studies, GIS registries and investment-package specifications that no investor could browse, so its projects could not reach the people who might fund them.

Built it from design through code and deployment with AI coding agents: eleven investment packages and five superzones generated from the project’s own registries, package search and comparison, an investor matcher, enquiry, meeting and NDA forms, and a staff area with review, scheduled publication and rollback.

Deployed as a release candidate in Persian and English. It takes an investor from the first search to an NDA in one place; the Arabic edition is not finished yet. ghadir.emkan.world

People working on Iran’s territory, economy, networks, resources and investment kept that spatial data in separate files and systems. Nobody could search it, compare two provinces or share a view in one place.

Built a national spatial platform from design through deployment with AI coding agents: 34 ready-made views, a profile for each of the 31 provinces with side-by-side comparison, one search across territory, sector, layer, dataset and opportunity, and an admin console for importing and verifying data.

Released to production in August 2026 for signed-in users only, deny by default, so sensitive layers stay inside the organisation that owns them. map.emkan.world

The Doqaroun free zone, on the Iran–Afghanistan border, is Iran’s shortest route to Afghanistan’s import market and on to Central Asia. An investor still had no way to see its land, its zones or which projects were ready for money.

Built the zone’s investment-promotion portal and back office with AI coding agents: six starting paths (find an opportunity, land and warehouses, enter the Afghan market, find a partner, propose a project, book a visit); a smart map drawn from the official KMZ files with the 434-parcel land-use plan; eight zones with area, jobs and capital; action-ready opportunities with a readiness score and a source; market flows and incentives; a three-step expression of interest, and an API for enquiries, investor accounts, consent records and document requests.

Online, in Persian and English, with every figure carrying its period and source; the zone’s official launch waits on its organisation’s approval. investindoqaroun.emkan.world

“A multi-trillion-dollar economy” stays a slogan until it says by which measure, by when, and under which assumptions about sanctions, war, peace, investment and productivity. Iran’s long-range vision needed numbers a reader can trace and scenarios that do not pose as forecasts.

Built an economic-intelligence product from design through deployment with AI coding agents: sourced indicators for Iran; three adjustable paths for 2050 and 2065; a benchmark of how 197 countries wrote their national visions, across 175 sub-questions and 40,627 comparisons; a data-quality dashboard that shows how much is directly evidenced; and four ways in: for citizens, for managers’ decisions, for investment, and a global atlas in English.

Deployed in August 2026. It turns a national promise into paths that can be calculated and criticised: every projection is labelled a scenario, a missing figure is stored as a coded gap, never a zero, and each source file carries its hash. visions.emkan.world

Iran sits between seas, markets and land routes, yet position alone brings no cargo and no capital. Policymakers, investors and researchers deciding on a transit corridor needed its routes, the ports and borders on them and the treaties that govern them in one place, with a source behind every claim.

Built the atlas of the world’s transit corridors, with Iran as the reference point, from design through deployment with AI coding agents: 135 corridor networks broken into 437 routes, 2,461 segments and 1,058 nodes; 234 legal records with 11,768 treaty–country links; 178 economic-geography points inside Iran; side-by-side comparison of up to five corridors; a stable page for each of 2,133 searchable assets; and a data room of 68 sheets and 37,966 rows that every figure traces back to.

Live in Persian and English. It answers the question a transit decision turns on: which routes cross Iran (56 records) and which go around it (77), and which port, border crossing or rule would change that. It keeps fact, inference and hypothesis visibly apart, and reads each corridor against 15 economic sectors. corridors.emkan.world

Territorial development is discussed in language no product team, no investor and no citizen can act on.

Built the ecosystem's information and content architecture — development missions, economic zones, sectors, library, gallery, podcast — so that policy becomes navigable.

In active development as Emkan's flagship platform.

An investment-promotion agency with no digital surface is a phone number. Investors need discovery, evidence, a way in, and a reason to stay.

Architected the full investor journey — opportunity discovery, competitive-advantage assessment, expression of interest, organisational engagement, request management and aftercare.

In active development as the digital infrastructure of Emkan's IPA.

Makran is one of Iran's largest development bets and almost nobody — investor, citizen or policymaker — can picture it.

Benchmarked 38 countries, ran 8 regional deep-dives, contributed to two documentaries and designed the region's investment-conference framework.

Fed directly into national coastal-settlement planning via the Economic Working Group.

Registering a company in Iran requires an expert because the process is deliberately opaque. The expertise is the product — and the bottleneck.

Modelled the specialists' tacit knowledge into a guided, question-based workflow that generated the filings.

Ran for a year. It taught me that productising expert knowledge is a knowledge-extraction problem long before it is a software problem.