samsullivan

Cardog | Toronto, Ontario | ONSITE, HYBRID, REMOTE (role-dependent) | Full-time | https://cardog.app/careers

Cardog turns raw vehicle data into facts you can check: VIN decode, market price, recalls, and what the car's own modules report. Insurers, dealers, manufacturers and the public bodies that register vehicles all use it. We're a small team and every engineer owns a whole area.

Data - Data Platform Engineer (Remote). Own the datasets: government files, manufacturer filings, dealer listings, signals from the vehicle. Decide which source wins when two disagree, and be able to prove it. SQL first, entity resolution, tens of millions of rows. Postgres, TypeScript, Workers, R2. - VIN Decode Engineer (Remote). Own decode accuracy. Read manufacturer filings line by line, write the patterns, and test them against millions of real VINs. For people who are obsessive about correctness. TypeScript, Postgres.

Design - Design Engineer (Remote). Own the component library and the screens built on it. Dense data, and making uncertain or missing values look uncertain. You design in the browser and you merge it yourself. React, Next.js, Tailwind. - Founding Engineer, Dealer Platform (Toronto / Remote). First engineer on software for dealer groups. Time in stores, then ship it the same week. Strong Postgres, multi-tenant.

Business - Account Executive (Toronto / Remote). Enterprise and public-sector sales cycles measured in quarters. Run technical first calls alone and pull your own numbers in SQL. - Policy & Standards Lead (Ontario). Represent us in the consultations and standards bodies writing the rules for vehicle data. Closer to counsel than comms.

Field - Hardware Engineer, Vehicle Interface (Toronto). Own the diagnostic-port device: board, firmware, enclosure, from bench to a unit on a dealer lot. C or Rust, CAN at the frame level, UDS/ISO-TP against real cars. - Forward Deployed Engineer (Ontario). Take named accounts from the first meeting to a running system. TypeScript, SQL, and some time under dashboards.

Apply on the page (each role takes applications in place) or write to careers@cardog.app.

Should credit the authors of the Verifiable Credential library. The vendor isnt the one to credit here.

Digital verification is going to matter a lot more for objects we own rather than the objects that proxy for that (currently the main function of an ID). Identity fraud is only problematic because ownership is tied to a loose record of SIN/DL.

Having a physical medium represent ownership just shifts the burden to the state and allows for social engineering and fraud to persist.

Telemetry is usually a standalone module. Heres one off a 26 Hyundai Palisade -- Telematics (DCU) Part num: 96510P9100 firmware (OTA updatable) 01F.7.CAN.3.000.18.714.9

Its only job is to detect harsh acceleration and brake events. Probably being sold to your insurer.

Two different things are getting called "car data" here.

Facts about the car: VIN, spec, recall status, odometer. Attested by someone other than the owner. Outlives every owner and the owner is the last to have it, if at all.

Facts about the driver: speed, location, timestamp. Thats what GM sold, the fix is to not collect it but OEMs seem to take 'anonymization' approach.

The DRIVER act treats both as the same which is why it fails to fix anything. The second needs a ban. The first needs the opposite - especially as we take the driver out of the vehicle - an authoritative record of the vehicle. How do we expect AVs to have adoption when the only safety certification is the company's press release?

Your mileage went from VW->telematics vendor->broker->carfax, thats the chain it usually follows. The fuse stops it but the real issue is the owner is the only one without the record.

You could probably find some interesting info if you send out some privacy requests through the chain. LexisNexis and Verisk are consumer reporting agencies, so they have to provide you a disclosure on request. Start there.

Cardog | Toronto, Canada or REMOTE | Full-time https://cardog.app/careers

We're the system of record for the Canadian vehicle: raw automotive data turned into facts that can be checked. Our open-source VIN decoder sits underneath a government vehicle registry.

Some of what's interesting:

- A government registry, a manufacturer's own filing, a listing site and an insurance record will describe the same truck four different ways, and nothing in any of those files says they're the same truck. Reconciling that, for every vehicle in the country, in a way that holds up when a regulator asks how you know, is most of the data work.

- Seventeen characters on a door frame encode manufacturer, plant, model year and build attributes — badly, inconsistently, and differently per manufacturer. Public reference data covers some of it. The rest gets worked out from filings, recall documents, and patterns inferred across millions of vehicles.

- We read vehicles directly over OBD-II now. Forty modules answered on a recent Hyundai, each reporting its own part number and installed software version. Nothing downstream of the VIN moves when a module is reflashed, so no registration, history report or dealer system knows what software is actually in a car.

Hiring: VIN Decode Engineer, Data Platform Engineer, Infrastructure Engineer, Founding Engineer (dealer platform). TypeScript, Postgres, Cloudflare Workers.

Full descriptions and apply: https://cardog.app/careers

Cardog | Toronto, ON | Remote/Hybrid (Canada) | Full-time AI-powered automotive platform — we’re building the infrastructure layer for vehicle data, from VIN decoding to market intelligence, powering government and public-sector vehicle programs.

Looking for: Automotive experience, Full-Stack, AI/ML, Data Engineers (public-sector/govtech experience a plus)

Stack: React Native/Expo, Next.js, Hono, Cloudflare Workers, PostgreSQL, TypeScript everywhere

https://cardog.app/careers

Cardog | Toronto, ON | Remote (Canada) | Full-time AI-powered automotive platform — we're building the infrastructure layer for vehicle data, from VIN decoding to market intelligence.

Looking for: Automotive experience, Full-Stack, AI/ML, Data Engineers

Stack: React Native/Expo, Next.js, Hono, Cloudflare Workers, PostgreSQL, TypeScript everywhere

https://cardog.app/careers

Cardog | Toronto, ON | Remote (Canada) | Full-time

AI-powered automotive platform — we're building the infrastructure layer for vehicle data, from VIN decoding to market intelligence.

Looking for: iOS/Mobile, Full-Stack, AI/ML, Data Engineers

Stack: React Native/Expo, Next.js, Hono, Cloudflare Workers, PostgreSQL, TypeScript everywhere

https://cardog.app/careers

not sure what PocketOS does or why your whole dataset would be a single volume without a clear separation between application and automotive data. how are you decoding VINs?

Cardog | Toronto | Full Stack Engineer | In-person

We build automotive software. We work with dealerships on ad-hoc projects, websites, and pricing tools. We also run consumer facing products - a web app for vehicle shopping, research and market data, and a mobile app.

Right now we're building EV battery health diagnostics - connecting to a car's BMS over CAN bus, pulling cell voltages and state of health, generating reports. Mix of hardware and software.

Small company with real clients and real revenue. One engineer. Looking for a second who wants to own things e2e.

hello[at]cardog{dot}app

The need for proxies in any legitimate context became obsolete with starlink being so widespread. Throw up a few terminals and you have about 500-2k cgnat IP addresses to do whatever you like.

why do we insist on bringing the worst tech from sci-fi to reality? is our collective intelligence still not enough to thwart the social and market dynamics?

all of these problems are better articulated at the level you just explained them. the code for these issues is convoluted and is only of use when an entity (human or not) can actually manipulate the symbolic text that achieves that task. a random oauth stub is of 0 use to the most skilled programmers without documentation as to what contracts and invariants are. bits in a file is just a means

Trying to make anything car related easier - Cardog.app

Buying, researching and analyzing automotive data is broken. Trying to fix that bit by bit

Imagine a world where the code we write for humans would actually integrate with other computers

The battery uncertainty is real, but I think the bigger issue is information asymmetry.

Looking at actual market data, the spread on used EVs is wild - a 2022 Tesla Model S ranges from $57 to $112k depending on trim/condition (https://cardog.app/tools/valuation/tesla/model_s/2022). That's a $60k spread on the same year vehicle. Compare that to ICE vehicles where the range is typically much tighter.

When buyers can't confidently price an asset, they discount heavily. The depreciation problem might actually be a data problem - we just don't have the standardized battery health reporting and historical comps that exist for ICE vehicles yet.

this is a claude output verbatim, wonder what happens to reddit over the next 3-6 months, we really destroyed the commons.

answering correctly is completely dependent on the attention blocks to somehow capture the single letter nuance given word tokenization constraints. does the attention block in kimi have a more receptive architecture to this?

Thats why they need to pay 300k for a slide designer https://openai.com/careers/creative-lead-presentation-design...

"figure out how your employer makes money and position your ass directly in-between the corporate bank account and your customers' credit card information."

we're going to learn why companies were invented the hard way this decade aren't we

Im building https://cardog.app out of Toronto

Trying to make car research, buying and ownership easy

MCP feels overengineered for a client api lib transport to llms and underengineered for what ai applications actually need. Still confuses the hell out of me but I can see the value in some cases. Falls apart in any full stack app.

But meeting at some random parking lot without knowing what you're actually buying seems to be the only option. Dealers charge a huge premium for the trust layer, just like hotels did before Airbnb. It's an adversarial market that desperately needs a trusted intermediary, but current players exploit this information asymmetry rather than solve it.

Cardog guides users through the complete vehicle ownership cycle. Start by researching models using our comprehensive specs database, owner reviews, and curated video content. When ready to buy, analyze real-time listings from sellers across North America, with data organized by exact trim and feature configurations.

After purchase, the garage becomes your ownership hub - tracking maintenance, storing documents, and monitoring market values. Get proactive alerts for service needs, recalls, and registration renewals. When it's time to sell, use historical price data to time the market and determine optimal listing price.

Unlike search engines providing scattered information, Cardog unifies vehicle data with practical ownership tools. Whether you're researching your next car or managing your current one, everything stays organized and updated automatically. The AI interface makes this comprehensive dataset accessible, while the garage helps track your vehicles through their entire lifecycle.

Each phase feeds into the next: research informs purchase decisions, purchase details populate your garage, and garage history helps optimize your eventual sale - creating a seamless ownership experience.

The LLM interface isn't the main feature - it's the combination of vehicle-specific data sources and tools:

Vehicle maintenance history aggregation Automatic quote retrieval from sellers Registration/insurance document management Market value tracking across North America VIN-specific recall and service bulletin monitoring Manual parsing and model-specific guidance

The LLM makes this data accessible, but the value is in having everything about your vehicle centralized and contextualized. Claude/Deepseek can give general car advice, but they can't track your specific vehicle's history or provide personalized maintenance alerts.

Would you find value in having all your vehicle's data and documents in one place?

With all the news about scraping legality you'd think a multi billion dollar AI company would try to obfuscate their attempts.

You should consider focusing on intercepting network requests. Most if not all sites I scrape end up fetching data from some api. Like others have said, if you instead had the LLM create an ad hoc script for the scraping task and then use the feedback loop to continuously improve the outputs it would be really cool. I'd pay between $5 - $50 for each working output script.

We're going to need some sort of knowledge graph for all the different vehicle models, imagine how vague dealers will be when your looking on the lot. Wonder how crazy this will be once they start adding in-car ads.

It would have to in order to produce the outputs, our brains have crazy physics engines though, F1 drivers can simulate an entire race in their heads.