Fordonsbud — a vehicle platform that runs itself
Fordonsbud (fordonsbud.se) is a Swedish platform for vehicle valuation, bidding and dealer flows: a private individual submits a registration number, receives a valuation and lets vetted dealers bid openly. Nuvid built the entire platform from the ground up and automated its operation — the valuation model, bidding, SEO content and mailings run in scheduled jobs around the clock. The result is a platform that largely runs itself and that, in practice, the owner only needs to keep an eye on.
Quick overview
In short: Fordonsbud is a complete vehicle valuation and bidding platform that Nuvid built and automated. A bespoke valuation model, open bidding between vetted dealers and a large library of programmatic SEO content are driven by scheduled jobs in the cloud — the platform collects data, publishes and sends mailings around the clock without anyone having to sit at the keyboard.
How we built it:
- A bespoke vehicle platform on serverless edge infrastructure (serverless hosting and functions, a database, object storage, key-value storage and an AI model) with valuation, bidding and dealer accounts.
- A self-learning valuation model that collects final prices and market prices every night and responds instantly in the window when there is enough data — otherwise an AI agent finds comparable vehicles and saves the findings.
- Programmatic SEO/GEO: hundreds of pages for valuing and selling cars, plus an llms.txt and IndexNow batch, so the content shows up in both search and AI answers.
Impact: frees up an estimated 20 hours per week of manual valuation, content production, prospecting and mailings — work that would otherwise have required a full-time person.
The challenge
Selling a vehicle privately is slow and uncertain. The seller rarely knows what the vehicle is actually worth, and comparing bids from several dealers takes endless calls and emails back and forth. On the other side sit dealers and buyers who want the right objects but lack a smooth way to find and bid on them. The market was full of manual work on both sides.
The challenge for any platform meant to solve this is that it requires an enormous amount of ongoing work to keep it alive: vehicles must be valued, prices collected, content written to show up in search, leads followed up and dealers activated. Do all of that by hand and it quickly becomes a full-time job — and then the business model doesn't work. The platform therefore needed not just to be built, but to be built so that it largely runs itself.
What Nuvid built
Nuvid built the whole of Fordonsbud as a product of its own — not an integration between existing systems, but a platform from the ground up. At its core are three parts that connect:
- Vehicle valuation. The seller enters a registration number, which is looked up automatically against the vehicle register, along with mileage and condition. A bespoke valuation model responds instantly in the window when there is enough data; otherwise an AI agent is set to work, finding comparable vehicles on the web, emailing the valuation and saving its findings so the model grows with every request.
- Open bidding. The request goes out to vetted dealers and buyers who place bids openly in the platform. Bidders register and verify themselves through their own flows, and auctions close automatically at the set time. Free for the seller — the buyer pays a fee when a deal goes through.
- Dealer flows. Dealers get accounts, onboarding sequences, inventory feeds synced to biddable objects and widgets for their own sites. Around this sit financing and insurance comparisons via partners, plus featured subscriptions with automatic invoicing.
Everything is built for many vehicle types — car, motorhome, caravan, motorcycle, boat, van, truck, construction machinery and more — and the same valuation and bidding logic is reused across the entire range.
How it's built — APIs, automations and infrastructure
Fordonsbud runs entirely on bespoke, serverless edge infrastructure, which means there is no server to manage and operation is essentially self-sustaining. The technical building blocks:
- Infrastructure. Serverless hosting (the web), serverless functions (logic and scheduled jobs), a database for leads, valuations, bids and dealers, object storage (image and vehicle objects), key-value storage and an AI model for AI-generated drafts. No server operations of our own — the platform scales and sustains itself.
- APIs. Around forty bespoke endpoints (serverless functions) drive valuation, bidder registration and verification, leads, feeds, widgets, auth with two-factor, payment and admin. Vehicle lookups happen automatically against the vehicle register via registration number.
- Automations and cron jobs. A separate cron worker with around thirty scheduled triggers pings internal endpoints around the clock: nightly collection of auction final prices and market prices into the valuation model, syncing of dealers' inventory feeds to biddable objects, automatic closing of auctions, prospecting, outreach, newsletters, article drafts, onboarding drip and monthly invoicing. It is this web of jobs that lets the platform "run itself".
- SEO and GEO — the platform's growth engine. Programmatic SEO at scale — entire page families for valuing vehicles (value-car, value-motorhome, value-caravan, value-boat) and hundreds of guides for selling a car in every conceivable situation. A daily SEO release and an IndexNow batch keep the content indexed, and an llms.txt describes the platform so it is cited correctly in AI search. This content ranks in search and gets cited by language models, which brings in visitors — and therefore sellers and dealers — organically, without paid advertising. The traffic is fundamentally an effect of the content, not an ad budget.
- Design. A bespoke, consistent frontend with clear flows from registration number to bid, built to feel safe and simple for both sellers and dealers.
- Third-party connections. A bot protection to stop bots without disrupting real users, Resend for transactional and notification emails, plus affiliate connections (Zmarta/Axo) for financing and insurance comparisons.
See more examples of similar work under our solutions.
How much time it frees up
The point of Fordonsbud is not just that it exists, but that it works without daily manual involvement. The tasks that would otherwise need a person on hand — valuing vehicles, collecting prices, writing content, prospecting dealers and sending mailings — are handled by scheduled jobs that run day and night. In practice the owner only needs to keep an eye on the platform rather than run it by hand. The time freed up is really only half the value: because the same automations also publish content and bring in traffic organically, the number of leads and deals grows without the workload growing with it.
| Manual task | How we solved it | Time freed up |
|---|---|---|
| Valuing every incoming vehicle by hand | Bespoke valuation model responds instantly; AI agent finds the rest and teaches the model | Most of the valuation work |
| Collecting final prices and market prices manually | Nightly cron jobs fetch and update the price data automatically | Several hours per week |
| Writing SEO content page by page | Programmatic page families plus AI-generated article drafts and a daily SEO release | A large share of the content work |
| Finding and contacting dealers | Automatic prospecting and outreach with daily quotas in scheduled jobs | Ongoing sales time |
| Following up leads and sending newsletters | Nudge, nurture and newsletter jobs run on a schedule | Recurring admin time |
All told, we estimate the automation frees up in the order of 20 hours per week compared with running the same platform manually. This is an estimate based on how many tasks are automated and how often they run, not an exactly measured figure — but the order of magnitude corresponds roughly to a full-time role that would otherwise have been needed for valuation, content, prospecting and mailings.
Results and next steps
The result is a vehicle platform that sustains itself: it collects data, improves its own valuation model, publishes content and drives contact with both sellers and dealers around the clock. The real business value is not just time saved, but that the platform brings in customers on its own — the programmatic SEO and GEO content ranks in search and is cited in AI answers, so new sellers and dealers find their way in organically without paid advertising. At the same time the cloud-based infrastructure means there are no server operations to maintain. This delivers more leads and deals without the cost or workload growing in step.
The next steps are about continuing to broaden the valuation data, connecting more vehicle types and dealer segments and refining the automated flows over time. If you want to build something similar — a product or platform that runs itself rather than another manual process — get in touch and we'll look at it together.
Want to see how this could look for you? Explore our solutions, see more cases and platforms we've built, or get three free suggestions.