How Jev works
- Published
- Author
Lucas Dworakowski
In plain terms: what TypeSafe’s System One model is, and why it returns probabilities instead of text.
ReadFrom projects, conferences, and workshops — IoT, AI, mobile, and problems that show up in real work.

In plain terms: what TypeSafe’s System One model is, and why it returns probabilities instead of text.
Read
In plain terms: what Model Context Protocol is, and why AI needs a shared way to reach external systems.
Read
Why higher MQTT QoS is not always more reliable in industrial telemetry, and how to model flow control.
Read
Attention, Q/K/V, and Transformer layers — how an LLM works on top of Google’s paper.
Read
How image classification and detection work: features, CNNs, softmax, and decision thresholds in production.
Read
From scan to minutiae: ridge extraction, descriptors, matching scores, and biometric decision thresholds.
Read
Pupil/iris segmentation, polar normalization, Daugman iris codes, and Hamming-distance matching.
Read
How an APNR/ANPR pipeline works: plate localization, matched filters, template OCR, and temporal fusion.
Read
Offline vs online signature verification: temporal features, DTW, embeddings, and decision thresholds.
Read
Public and private keys, RSA in short, digital signatures, key exchange, and common implementation pitfalls.
Read
Syntax, type safety, interoperability, memory, performance, and when Objective-C still shows up in iOS projects.
Read
Retain, release, weak/unowned, reference cycles, and what the compiler inserts for you in Objective-C and Swift.
Read
Pros and cons of native development vs React Native: UX, performance, cost, team shape, and practical decision scenarios.
ReadNo pitch. Pick a day and time — we’ll confirm by email, or suggest another slot if the calendar is already taken.
Rather write? Contact form
October 2026
Time zone: Europe/Warsaw
Time