Kandua promises to simplify home repairs by eliminating the need for homeowners to play detective or chase tradespeople across multiple apps. Jess, the platform’s AI companion, aims to translate vague complaints into clear job briefs. The real question is whether this actually reduces the stress of getting a geyser fixed, a breaker sorted, or a bathroom patched, or if it just repackages the existing mess.
The difference matters because home repairs consistently fail in predictable ways. The wrong person arrives, the quote changes, the job is left unfinished, or payment becomes awkward. A platform claiming to reduce this friction must demonstrate control, not just speed.
Jess should help you describe the problem, not pretend to solve it
First, we need to determine if Jess acts like a helpful intake form or a chatty sales script with a friendly avatar. For a leaking geyser, a confident diagnosis is not useful. Instead, a series of pointed questions are needed: where the water comes from, if it’s hot or cold, how long it’s been leaking, and if there’s rust or visible damage. This level of detail helps a plumber decide if the issue is a valve, a tank, a fitting, or something more serious.
The same logic applies to tripping lights. A good prompt doesn’t guess at an electrical fault from one sentence. It asks which breaker trips, if it happens immediately or after a while, if a kettle, heater, or tumble dryer triggers it, and how often it repeats. This is guidance, not diagnosis. A chatbot that acts certain about an unseen electrical problem is reckless, not helpful.
Cracked bathroom grout sounds minor until you ask why it happened and how far it has spread. Jess should treat this as a cue for clarification, not a chance to improvise. The right questions are whether the cracking is cosmetic, if water has gotten behind tiles, if the area has loosened, and if the problem requires a tiler, a plumber, or a proper inspection before patching begins.
This distinction is crucial. A platform earns trust by narrowing the job sufficiently for a qualified person to quote accurately, then stopping its attempts to sound like the qualified person.
The platform has to earn the handoff
Kandua aims to be more than a directory with a chat box. Its pitch is that it manages the job from initial description through booking, payment, and follow-up, instead of simply selling your details. This aspect is worth examining, as many trade platforms are essentially lead farms disguised as convenience.
If the system works, the user should describe the problem once, get matched with the right professional, and complete the process without turning it into a group project. A geyser leak should go to a plumber. Repeated tripping should go to an electrician. Cracked grout should go to a bathroom specialist, not a general handyman with vague confidence.
The promise is not magic; it is compression. Fewer calls, fewer screenshots, fewer “send me your address again” messages. It reduces the usual chaos of coordinating a repair across Google results, missed calls, and unverified Facebook recommendations.
Google and Facebook still win on volume, but not on control
Searching Google for a plumber in Durban North or a sparky in Fourways will provide options. However, you bear the full burden of proof. You must determine who is legitimate, available, insured, and if the five-star reviews are genuine. This often involves cross-checking business names, calling multiple numbers, and inferring professionalism from a WhatsApp profile picture.
Facebook groups are worse in a uniquely South African way. They can be excellent when someone posts a photo of a leaking ceiling and neighbors offer recommendations. They can also become rumor mills. One person had a good experience three years ago, another a disaster last month, and neither story confirms if the person arriving at your gate has the right qualifications.
Kandua’s value proposition is not that it finds people you couldn’t find yourself. It tries to sort the mess before it reaches your driveway. Background checks, qualification verification, and a platform that keeps transactions within its own system all aim to reduce guesswork.
The quote process is where the promises get exposed
A repair platform’s success hinges on the quote. If it’s vague, late, or obviously padded, the entire system collapses into the same old problem with a shinier interface.
The ideal Kandua should provide multiple quotes from vetted professionals, often within hours or a day. This allows homeowners to compare not just prices, but also availability and profile information. A single quote is just a number; three to five comparable quotes reveal if someone is overcharging, underbidding, or pitching for work they don’t truly want.
This comparison also reflects the platform’s quality control. If a plumber quotes the geyser leak, an electrician the breaker issue, and a bathroom specialist the grout repair, the matching is effective. If recommendations are sloppy, the AI is merely decorative.
The user experience should also accommodate the necessary but mundane aspects of hiring someone: arrival time, job duration, and what happens if the issue is larger than initially estimated. These questions consume time and patience when dealing with contractors individually. A good platform answers them before they become arguments.
The payment system is the part most people only think about after something goes wrong
Kandua’s claim becomes more significant here. If the platform handles payment through its own system, rather than having you send money directly to a stranger, it removes a major risk in home repairs. The ideal setup is straightforward: the job is booked, the work is done, the invoice is issued through the platform, and payment is released once the homeowner marks the job complete.
This structure doesn’t eliminate disputes, but it changes their nature. You’re not chasing someone who vanished with a deposit. You’re dealing with a platform that has records of the quote, chat, appointment, and invoice. If the final amount exceeds the original estimate, the platform should facilitate a proper conversation before additional work proceeds. No surprise invoices or “we had to do more” explanations after payment has already moved.
Here is the practical test I would care about most:
| Situation | What you want | What to watch for |
|---|---|---|
| Geyser leak | Clear symptom questions, plumber matching, written quote | Whether the platform treats a leak as a guessable fault |
| Electrical tripping | Safety-first triage, electrician referral, inspection advised | Whether Jess pretends to diagnose a fault it cannot see |
| Cracked grout | Correct trade category, scope clarified, quote before work | Whether the platform recognises when water damage may be involved |
Financial risk extends beyond the invoice; it includes who bears the inconvenience when things go wrong. A platform with proper payment handling should reduce the chances of paying upfront for poor work, no-shows, or unfinished jobs.
Guarantees, cancellations and damage complaints decide whether the platform is serious
Many service marketplaces make grand promises until someone asks about responsibility when work goes awry. Kandua should be judged on whether it makes the professional’s own warranty visible and usable, not on a broad blanket promise it cannot realistically honor across every trade.
The same applies to cancellations. What is the policy if a job is canceled last minute? Who pays the fee? How much notice is sufficient? Homeowners often feel powerless in these arrangements, and contractors are also used to being left hanging. A trustworthy platform must define the rules before irritation begins.
Then there’s damage. If a wall is cracked, a fitting breaks, or a job leaves a mess, there needs to be a complaint process beyond a neglected customer service email address. Evidence is key here, as the platform has job history, messages, and user-uploaded photos. This provides a better chance of mediation than the typical “please call the person directly” dead end.
The real product is less panic, not artificial intelligence
The smart way to evaluate Kandua is to disregard the AI hype and focus on the operational burden it removes. If Jess helps homeowners clearly explain problems, if the platform finds the right tradesperson, if quotes are comparable, if payment is controlled, and if disputes have a clear resolution path, then the product is useful.
Conversely, if Jess speaks as if it knows the fault when it only knows the symptoms, the entire system becomes another layer of confident nonsense over an already stressful task.
Home repairs are difficult not because people cannot find plumbers or electricians. They are difficult because trust is expensive, time is scarce, and bad information is pervasive. Kandua’s pitch is that it reduces all three. That is a proper claim, and it deserves a proper test.
