Use casesOperations & logistics
Let warehouse staff ask 'how do we do X here?' and get the SOP answer
Staff ask 'how do we pack lithium batteries?' and get the answer from your own SOPs, with a citation to the exact section — in their own language.
- How it works
- Your standard operating procedures, packing instructions, carrier cut-off rules and safety sheets are indexed with embeddings. When someone asks a question, the closest passages are retrieved and the small model answers in plain language, citing the exact document section: 'Lithium batteries go in UN3481-marked boxes, max 2 per parcel — see Packing SOP §4'. Staff can ask in their own language even if the SOPs are written in another — useful with international temporary workers.
- Data you need
- Written SOPs as machine-readable text: Word documents, wiki pages, plain-text exports. This is the honest gate, twice over — many SMB warehouses have procedures only in people's heads, and much of what is written exists only as scanned PDFs or laminated printouts. Scans must be exported or typed up first, as no vision model is served; if nothing is written at all, write the documents first, because the system cannot retrieve what does not exist.
- What to expect
- Good at answering questions the documents actually cover, and the citation makes every answer checkable. Answers are only as current as the documents — an outdated SOP produces confidently outdated answers — and questions outside the documents should be refused rather than improvised, which is how the system is configured. Staff also need a device on the warehouse floor to ask from; without one, the shift lead stays the interface.
- Where people stay involved
- Every answer shows its source passage so staff can verify. For safety-critical procedures (dangerous goods, machinery) the answer should point to the document rather than be treated as the authority, and the shift lead remains the escalation path. Someone must own keeping the index in sync with document changes.
Which model, and what it costs to run
Qwen3-32B
This job answers while someone waits, so latency comes first: we hold it to a single small card and, within that, take the best independent score on structured output and tool calls.
- Licence
- Apache-2.0
- Weights at 4-bit
- 18 GB
- Context
- 32K tokens
- Publisher
- Alibaba (Qwen Team)
What the hardware costs
One 48 GB card holds it
- Rent in the EU
- $1.60/hrScaleway, Paris (PAR2)
- Buy the card
- $7,569new, one-off
- Or rent it by the token
- $0.12 / $0.12per M in / out · Nebius AI Studio · EU
Hardware only, third-party prices from 2026-07. The figure excludes the KV cache, which grows with context length and how many people use it at once — sized properly in a conversation, not guessed here. Renting by the token is cheaper up front; why our customers still self-host is below.
Structured output and tool calls
Measured in a sandbox, on somebody else's functions. It tells you which models are capable of the shape of the job, not which one will survive contact with your API.
Independent measurement · UC Berkeley (Gorilla project) · board updated 2026-04-12
The API is cheaper per token. Here is why our customers don't use it.
We will not pretend otherwise: renting a model by the token from a serverless API costs less per million tokens than a card we run for you. We show that price on every use-case page. What it does not include is the part a shop with a customer database actually pays for.
- 01
Your data never leaves hardware you can point at
A serverless "we don't retain your data" is a clause in a contract. Running the model on a card in Amsterdam is a fact of architecture: your catalogue, tickets and customer records are never sent to a third party at all. For a GDPR audit, that is the difference between a promise and a floor plan.
- 02
The price cannot move without your say-so
A serverless rate card is somebody else's lever. The provider can raise the price, retire the model, or change the terms, and your cost moves with it. The same model on the same card costs the same next year — you own the number.
- 03
The model cannot be taken away
Hosted APIs deprecate models on their own schedule; the one you built on can be gone in a quarter. An open-weight model on your own hardware runs for as long as you keep the lights on. No vendor can end-of-life it out from under you.
And the price gap closes with volume: past a card you keep busy — very roughly four billion tokens a month — owning is cheaper outright, even before the three reasons above.
Getting a case like this one from a conversation to production takes about two months, and you can stop at the end of any phase.
Also in operations & logistics
All 13 →Sort supplier and carrier emails into the right queue
Tags every email to your purchasing or logistics inbox — delay, invoice, complaint — and routes it to the right queue, on your own infrastructure.
Pull order lines out of supplier PDFs and emails into a spreadsheet
Extracts SKU, quantity, price and delivery date from supplier PDFs and emails into rows you can import into your ERP or a spreadsheet.
Check supplier order confirmations against what you actually ordered
Compares each supplier confirmation against your purchase order and flags changed prices, quantities, substituted items and pushed delivery dates.
A morning digest of every delivery problem, in one email
One plain-language email each morning: the day's delayed, failed and held shipments, and the customers and orders behind each tracking number.
Next
Tell us what your team does by hand
Describe the process that takes the most time. We will say plainly whether a model is the right tool for it.