15-day free trialStart XGEN right in your browser — nothing to installStart free trial

Solutions · Customer cases

IT & Manufacturing customer cases

Customer cases applying AI to development and engineering — AI coding assistants and technical documentation

01Overview

AI for it & manufacturing work

IT and manufacturing organizations can rarely send source code or design and technical documents outside, and their development environments are often air-gapped. Plateer AI Labs applies AI here by building AI coding assistants and technical-document search inside the customer's environment, connected to in-house LLMs and dev tools.

Products and capabilities used

02Challenges

What it & manufacturing teams run into

  1. 01

    External coding AI is off-limits

    Concerns about source-code leakage often restrict commercial cloud coding tools.

  2. 02

    Technical documents are scattered

    Design specs, API docs and incident history sit in different repositories, so finding information takes time.

  3. 03

    Company-wide rollout needs operating rules

    Expanding a team pilot company-wide requires user management, usage tracking and model-selection rules.

03Approach

How it is applied, step by step

  1. 1

    Connecting in-house LLMs

    The coding assistant connects to LLMs running on the customer's GPUs so code does not leave the environment.

  2. 2

    Used inside dev tools

    It helps write, explain and review code inside the IDE, and DexOn lets developers move between vibe coding and hands-on coding.

  3. 3

    Indexing technical documents

    Design, API and operations documents are combined with RAG to answer questions with sources.

  4. 4

    Phased rollout

    After confirming usage patterns in a pilot group, scope expands with defined permission and usage rules.

04Deployments

What we have delivered

A large semiconductor manufacturer|Rolling outAI Coding Assistant

An AI coding assistant, rolled out company-wide

An AI coding assistant in a semiconductor manufacturer's internal development environment, connected to a local LLM and rolled out company-wide

Challenge

External coding assistants were not usable for security reasons, and answer quality for Korean-language technical questions had to be raised inside the internal environment.

What we built

An AI coding assistant was built against the internal local LLM, with Korean answer quality improved. Training ran alongside, spreading real usage across the company.

Scope delivered

  • Connected to an internal local LLM
  • Korean answer quality improved
  • Company-wide rollout with training
Read the full case
A defense and systems company|DeliveredAI Coding AssistantXGEN

An XGEN-based AI coding assistant in an air-gapped network

An XGEN-based AI coding assistant deployed on-premise into an air-gapped development environment with strict security requirements

Challenge

An air-gapped environment with no external connectivity still needed a coding assistant to raise development productivity.

What we built

An XGEN-based AI coding assistant was deployed on-premise, providing code support entirely inside the closed network.

Scope delivered

  • On-premise, air-gapped deployment
  • XGEN-based AI coding assistant
Read the full case

05Typical work

  • 01Code writing and explanation in air-gapped environments
  • 02Q&A over design and API documents
  • 03Drafting code reviews and test code

06Before you adopt

  • Suggestion quality depends on the connected LLM and context settings.
  • Generated code should go through existing review and test procedures.
  • Concurrent users and response speed vary with GPU capacity.
IT & Manufacturing customer cases · Plateer AI Labs