Questions, answered plainly.

Catalyst Gen is a technology company that builds interpretation layers: products that take something people already do informally and turn it into structure a system can act on.

That is one idea applied three times. HEREIQ does it for an organisation’s workforce, turning attendance taken on paper and leave agreed over WhatsApp into a single record payroll can close from. CelerCard does it for money that moves perfectly well locally but that an online checkout page cannot read. AI Intent Layer does it for a request typed to an AI model, turning plain language into an instruction the model can act on.

Catalyst Gen is a product company, not an agency. It does not take on client projects, build software to order, or bill for development time.

It identifies a problem worth solving, builds a focused product around it, runs that product in production, and improves it from what the people using it report. Migration, training and long-term support exist, but they are how a Catalyst Gen product lands inside an organisation — not a separate service line sold on its own.

Catalyst Gen was founded in The Gambia and serves customers worldwide.

Building there is a deliberate advantage rather than a constraint: it puts the company close to problems that remain underserved by mainstream software, and close enough to the people living them to know when a feature is theory. The first market for a product is usually The Gambia; the standard it is built to is not.

Catalyst Gen builds a small number of products deliberately, and runs each one end to end rather than shipping many and maintaining none.

Three are public today: HEREIQ, CelerCard and AI Intent Layer. A product only joins that list if it passes the same test — it has to take something informal and turn it into structure a system can act on.

HEREIQ is a workforce management platform that connects attendance, leave, payroll, employee records, documents, performance, onboarding and reporting into one system.

The problem it solves is reconciliation. When attendance lives on paper, leave is agreed over WhatsApp and payroll sits in a spreadsheet, those are three separate records that somebody has to match up by hand every month-end. HEREIQ makes them one record, so payroll closes from the same data attendance was taken in.

Yes. HEREIQ is the platform previously known as AttendanceGM, rebuilt around everything learned from running it.

AttendanceGM is the earlier name of the same product and the same team, which is why customer quotes and the product site still refer to it. It is a rename and a rebuild, not a different or replacement product, and existing AttendanceGM customers are HEREIQ customers.

HEREIQ is built for NGOs, country offices, and organisations operating across multiple locations.

It works offline, because the organisations using it do not always have a connection — attendance taken in a field office syncs when the connection returns rather than requiring one at the moment it is recorded.

Yes. HEREIQ is live and in production use with real organisations today.

It is the most established of the three products. The product site is https://www.gethereiq.com/; AttendanceGM is the earlier name of the same platform.

CelerCard is a virtual card designed to let people pay online without a bank card, funded from money they already hold.

The problem it addresses is a specific one: in mobile-first markets money moves well locally, and then a checkout page asks for a card — at which point the money someone has stops being money that page will take. CelerCard is being built so the card details work wherever cards are accepted online, with a physical card to follow that shares the same balance, controls and history.

CelerCard is in early access, opening in stages and starting in The Gambia. Places in the first group are limited.

Early access is a place in the queue for the first group, not an account. CelerCard does not hold funds, issue cards or process payments today.

AI Intent Layer is a tool that reads what someone is trying to accomplish as they write to an AI model, and turns their plain language into an instruction the model can act on.

Getting a useful result out of an AI model still depends on knowing how to structure a request, which is a skill almost nobody was taught. AI Intent Layer detects intent in real time, optimizes the prompt while you type, identifies context the model still needs, and adapts the result across ChatGPT, Claude and Gemini.

AI Intent Layer is in private beta, with a public waitlist.

Multi-language support, enterprise AI governance and workflow intelligence are on the roadmap rather than available now.

No. MediTicket, Nyaama and FireAlert GM are prototypes, not Catalyst Gen products, and none should be relied on in place of the service it models.

All three work and all three were built in days rather than quarters, mostly at hackathons. None is operated by, or affiliated with, the institution it was designed for. Where a real hospital or agency is named in one of them, it names the problem — not a partnership.

The prototypes exist because each one does at a weekend’s scale what a Catalyst Gen product does at a company’s: takes something people handle informally and turns it into structure a system can act on.

A hospital queue you join before leaving the house, a learning path explained in a language you think in, a fire report that already knows where it is. A project that cannot be described as that kind of pair does not get published.