Inference for faster, cheaper & smarter retail agents
Balance holiday peaks and shopper privacy with one API for open and fine-tuned models, serverless and in your own cloud.
- SigLIP so400m
- GLM-OCR
- Qwen3.8 27B
- Your fine-tuned model
- GLM-5.3 Flash
- DeepSeek V4.1 Flash
- Gemma 4 31B
The leading retail AI teams already run on open models
-
arxiv.orgSep 21, 2026
Graded-Relevance Composed Multimodal Retrieval for E-commerce Visual Search at Scale
…a PaliGemma2 bi-encoder trained on 3.5M graded pairs… The system is deployed in production at Walmart, where it’s serving live visual-search user traffic.
-
aws.amazon.comAug 28, 2026
How Decathlon runs demand forecasting at scale with Chronos-2
…Decathlon selected Chronos-2 as a core component of their forecasting stack… Every 6 months, a fine-tuning job built on AutoGluon adapts Chronos-2 to the latest data…
-
arxiv.orgSep 25, 2026
Retail Product Search: A Practical Approach at Target
We therefore chose e5-small-v2… as the base encoder for fine-tuning… The resulting system is deployed at scale and serves millions of guests daily.
-
mirakl.comMar 13, 2025
Mirakl sees 30% GMV growth and ARR of $177m as it doubles AI investment
By embedding AI across our suite of solutions and fine-tuning advanced models like Meta’s Llama, OpenAI’s ChatGPT, and Mistral’s 7B for commerce-specific use cases…
Keep up with the holiday rush, not your provider’s rate limit
Each job runs on a model sized for it, next to your app. When every store and shopper is busy at once, you add GPUs to your own cluster instead of waiting on a provider’s quota.
Run a month of catalog and store work for 94% less
Task-sized open models do each job for a fraction of a frontier API’s price. Each line is priced from its task page, at an illustrative month’s volume.
- Tag product photos by type, colour and material
$1,750$23 More attributes right - Read receipt photos for returns and rewards
$8,000$500 More photos read right - Route customer messages to the right team
$120$18 Labelled correctly: SIE 87.5% · GPT-6 93.8%
Tuned on your task, an open model beats the frontier APIs at it
How each result was measured
- Business questions to chart specs
- Superlinked’s gd-1, August 2026: Gemma 4 31B tuned on generated questions, then scored once on a BI assistant’s 875-question benchmark, counting the answers that return the reference data. GPT-5.6 Sol at maximum reasoning was the strongest API. The training questions came from the benchmark’s own generator, which is the home advantage tuning on your own task gives.
- Possible fraud in bank messages
- BANKING77: 3,076 customer messages to a bank, 240 of them reporting a payment they don’t recognise or a card or phone lost, stolen or compromised, none among the 9,993 the model was tuned on. The share of those reports caught while 95% of the messages flagged are reports, each model’s threshold set on the scored messages; GPT-6 Luna at low effort.
- Contract-clause review
- LexGLUE UNFAIR-ToS: 3,882 Terms of Service clauses labelled by lawyers, none of them among the 5,532 the model was tuned on. F1 on the unfair class; each API at its best setting, with the lawyers’ annotation rulebook in its prompt.
Meet the data requirements of your retail customers
| What the rule asks | SIE Serverless | SIE in your cloud |
|---|---|---|
Use shopper data only for the contracted purpose California CCPA regulations, §7051A service provider’s contract must “Prohibit the service provider or contractor from retaining, using, or disclosing the personal information that it collected pursuant to the written contract with the business for any purpose other than the business purpose(s) specified in the contract”. | Superlinked is a service provider Your contract with us sets the purpose | No service provider Shopper data stays in your cloud |
Never pass on face data without consent Illinois BIPA, §15(d)“No private entity in possession of a biometric identifier or biometric information may disclose, redisclose, or otherwise disseminate a person’s or a customer’s biometric identifier or biometric information unless” the person consents or a narrow exception applies. | Blur faces first Send shelf and product images, not face data | Faces stay with you Store images never leave your cloud |
Move EU shopper data abroad only under the GDPR EU GDPR Article 44A transfer to a third country may take place “only if, subject to the other provisions of this Regulation, the conditions laid down in this Chapter are complied with by the controller and processor”. | US or EU region Requests stay in the region you choose | Inside your network Your cloud account, in your region |
Delete or return shopper data at the end EU GDPR Article 28(3)(g)A processor’s contract must require that it, “at the choice of the controller, deletes or returns all the personal data to the controller after the end of the provision of services relating to processing”. | Agreed in your contract Deletion terms are part of the agreement | Nothing to return Shopper data never leaves your cloud |
Start on SIE Serverless, or run it under your control
SIE Serverless
We host the models; you call one API.
- No idle GPUs, pay for what you use
- Fits your stack: SDK, API, CLI, MCP
- Self-host the same open-source engine
- SOC 2 Type 2, US or EU data residency
- Create an account to try Serverless
SIE Cluster
Kubernetes in your own cloud, backed by our engineers.
- Terraform modules for your cloud
- Apache-2.0, same engine as Serverless
- GPU workers scale to zero between jobs
- Separate worker pools for each tenant
Deploy SIE to our AWS account with the superlinked/sie/aws Terraform module. Docs: superlinked.com/docs/deployment.Deploy SIE to our Google Cloud project with the superlinked/sie/google Terraform module. Docs: superlinked.com/docs/deployment.Deploy SIE to our Azure subscription with the superlinked/sie/azure Terraform module. Docs: superlinked.com/docs/deployment.Deploy SIE to our Alibaba Cloud account with the superlinked/sie/alicloud Terraform module. Docs: superlinked.com/docs/deployment. SIE Local
The open-source server on your own machine.
- Runs on your NVIDIA GPU
- One container, no cluster
- Process your data fully offline
- Same SDK and model IDs as Serverless
docker run --gpus all -p 8080:8080 ghcr.io/superlinked/sie-server:latest-cuda12-default Start from your task or your industry
Search and RAG
Documents
Text and LLM apps
- Chat Open LLM, OpenAI-style API
- Coding agents Tool calls for your coding agent
- Structured output Schema-valid JSON
- Classify text Your labels, no training
- Yes/no decisions Automate workflow checks
- Route requests Cheap or strong LLM, per request
- Extract names and dates People, orgs, places
- Knowledge graph Entities and their relations
- Translate text English into many languages