Data normalization and matching
Turning inconsistent source data into something you can query with confidence — deduplication, entity resolution, reconciling records that disagree with each other.
What I do
Most engagements start the same way: there is a dataset that should be clean and is not, or a pipeline that should be boring and is not.
Pipeline architecture
Ingestion from multiple upstream sources, scheduled refreshes, and the backfills you need when an upstream provider changes its mind.
API design and delivery
The service layer that puts processed data in front of your product, with the latency and reliability characteristics your customers assume.
Vehicle data systems
A long-standing specialty: VIN decoding, trim identification, and specification pipelines built on vPIC, EPA, NRCan, and related sources.
LLMs applied to real datasets
Using language models where they genuinely outperform rules — classification, extraction, disambiguation — with evaluation harnesses that prove whether they did.
AI-assisted engineering workflows
Getting a team's day-to-day development work onto AI tooling without losing review discipline or shipping code nobody understands.
Infrastructure and CI/CD
Build pipelines, deployment, database operations, and the monitoring that tells you something broke before a customer does.
Technical advisory
Second opinions on architecture, data strategy, and build-versus-buy, for teams who want the question answered rather than a deck about it.
Where this fits
- Automotive dataVehicle specifications, history, valuation, and the decoding layers underneath them.
- Marketplaces and listingsCatalogue quality, attribute extraction, and matching listings to canonical records.
- Data products and APIsCompanies whose product is the data itself, where accuracy is the feature.
- Teams adopting AI toolingEngineering groups deciding where models belong in their workflow and where they do not.
About
AltaMind Technologies is the consulting practice of Zhouwu Huang, a software engineer working on data systems, backend infrastructure, and AI-assisted development.
The work is mostly the unglamorous middle of a product: the ingestion, the matching, the corrections, the pipeline that has to run at 3am whether or not anyone is watching. Done properly, nobody thinks about it again — which is the point.
You work with me directly. There is no account manager, no bench, and no handoff to someone junior after the first call.
How engagements run
- Scoped per project, or a monthly retainer for ongoing work.
- Remote, with overlap for Pacific and Eastern time zones.
- Invoiced in Canadian or US dollars, whichever suits your accounts payable.
- Comfortable working under NDA and inside your existing repos and review process.
- Honest scoping: if the problem is smaller than an engagement, I will say so.
Start a conversation
Send a few sentences about what you have, what it is supposed to do, and where it stops doing it. I will tell you honestly whether I am the right person for it.