Insight · Plateer AI Labs

Article series

LLM Inside

The core ideas you need to make sense of an LLM, taken one at a time.

4 articles

The GS Certification Diary

Three weeks of reading 80 defect reports and rebuilding the code, the errors, and the manual

5 articles

The Harness Diary

Designing the execution environment an AI agent works in, from the ground up

10 articles

The Ontology Diary

Starting from the questions, building a knowledge graph, then working on quality and search

13 articles

All articles

Tech Note2026.09.04

For voice AI, operating criteria matter more than recognition accuracy

How we connected live speech input to XGEN, displayed interim recognition while the user spoke, and handed off a request only after the utterance was complete.

Somin Park

Tech NoteLLM Inside2026.09.03

If an LLM only predicts the next token, how does it get work done? (Part 1)

An LLM predicts the next token rather than looking up an answer. How that objective enables useful work—and why fluent output is not proof.

김해수

XGEN Preview2026.09.01

Deks — bringing enterprise AI agents into the actual desktop

The environment AI runs in and the environment people actually do their work in are separate. Deks is the desktop interface that connects XGEN AI agents to the user's PC, so real work gets done with the files and applications already there.

PPlateer AI Labs

XGEN Preview2026.08.28

Deks from install to first task — six steps

Install Deks, ask Geny in plain language to build an agent, connect that agent to your PC, and hand it real work. Six steps, with real screens.

PPlateer AI Labs

XGEN Preview2026.08.19

Geny — from an AI that answers to an AI agent that works

The XGEN agent takes another step forward. Geny is a new agent runtime with memory and a workspace of its own, able to find the tools it needs — and build them when they do not exist — to carry work forward.

PPlateer AI Labs

InsightInsight from our research and the field

The tools our agent built for itself were not canvas nodes

Article series

LLM Inside

The GS Certification Diary

The Harness Diary

The Ontology Diary

All articles

For voice AI, operating criteria matter more than recognition accuracy

If an LLM only predicts the next token, how does it get work done? (Part 1)

Deks — bringing enterprise AI agents into the actual desktop

Deks from install to first task — six steps

Geny — from an AI that answers to an AI agent that works