Skip to content

Case studies

Results, not promises

Every case follows the same format: context, problem, what we built, and the numbers before and after.

Erudil: an AI engine that prices every match outcome

Erudil (in-house product)

25k+

matches analysed

Erudil: an AI engine that prices every match outcome

Our own product. A probability engine turns match data into outcome probability matrices, compares them with bookmaker odds, and publishes a tamper-proof track record of every pick — wins and losses alike.

Sports analyticsPythonPyTorchPostgreSQL
AI knowledge assistants for a 30 000-employee grocery chain

Grocery retail chain, EU & US (under NDA)

−55%

time spent searching regulations

AI knowledge assistants for a 30 000-employee grocery chain

Three RAG assistants on one platform — for commercial staff, store directors (by voice, from the shop floor) and HR — answer questions from corporate regulations with a citation to the exact clause, or say honestly that they do not know — so staff stop phoning head office for routine questions.

RetailPythonLangChainPostgreSQL + pgvector
Document AI that reads 1 300 supplier invoices a day

Grocery retail chain, EU & US (under NDA)

88%

straight-through, no human touch

Document AI that reads 1 300 supplier invoices a day

Scans and photos of invoices, delivery notes and acts go through a multimodal model that extracts typed fields with a confidence score per field, reconciles them with the supplier and contract master data, and posts to the ERP. Only low-confidence documents reach a human — roughly one in nine.

Retail · FinancePythonPyTorchMultimodal LLM
Catalogue AI: classifying new SKUs and cleaning 182 000 product cards

Grocery retail chain, EU & US (under NDA)

−78%

time to classify a new SKU

Catalogue AI: classifying new SKUs and cleaning 182 000 product cards

Two models on the product catalogue of a grocery chain: one suggests the classification code and category for every new item with top-3 probabilities, the other finds duplicate product cards semantically. Category managers confirm with one click instead of searching the classifier by hand; every confirmation feeds back into training.

Retail · Master dataPythonscikit-learnSentence Transformers
Voice of customer: 12 review sources, one AI-classified stream

Grocery retail chain, EU & US (under NDA)

2.4 h

median reaction to a critical review, from 52 h

Voice of customer: 12 review sources, one AI-classified stream

Reviews from maps, app stores, social networks, the hotline and the in-store complaint book flow into one system that classifies each by topic, sentiment, store and severity, escalates critical ones to the responsible manager within minutes, and gives management a live, per-store picture instead of a monthly digest.

Retail · Customer experiencePythonTransformersPostgreSQL
KLR.gg: automated CS2 match analytics for private lobbies

KLR.gg (in-house product)

100%

of match days ingested with no manual upload

KLR.gg: automated CS2 match analytics for private lobbies

A platform that pulls server logs, match stats and demo recordings from Counter-Strike 2 community servers automatically, scores every duel and round, and turns them into player ratings, trends and balanced teams — no uploads, no spreadsheets.

Gaming · Esports analyticsPythonpandasFastAPI
Deep-learning MRI reconstruction: sharper images from 4× faster scans

Medical imaging research group (under NDA)

0.93

SSIM at 4× acceleration, from 0.71

Deep-learning MRI reconstruction: sharper images from 4× faster scans

A research-use system that reconstructs and denoises MRI series with neural networks — recovering image quality from undersampled and low-SNR acquisitions — deployed as a DICOM node between the scanner and PACS, with automatic quality gates so that no unreliable image reaches a reader.

Healthcare · Medical imagingPythonPyTorchMONAI

Discuss your project