Local AI IT Support Assistant

In Progress | Enterprise IT Environment

Identified an opportunity to reduce help desk ramp-up time by centralizing institutional knowledge into an

AI-powered support tool. Scoped and received leadership approval to build a local AI assistant that

serves as an on-demand resource for common IT support tasks.

The assistant uses a retrieval-augmented generation (RAG) architecture built on Ollama and

AnythingLLM, meaning it pulls answers directly from curated internal documentation rather than

generating responses from scratch.

Currently authoring the documentation that will serve as the models knowledge base, with deployment to

follow.

Tools: Ollama, AnythingLLM, RAG architecture

Status: Documentation in progress. Deployment pending.

Deployment Automation Script

In Progress | Enterprise IT Environment

Identified a multi-step software deployment process that required manual intervention at each stage,

creating inconsistency and technician dependency for routine installations.

Designed and built a PowerShell automation script to handle the full deployment sequence without

technician input at each step. The script executes the following stages automatically:

1. Creates and populates a local working folder

2. Installs the peripheral device driver

3. Verifies the device connection

4. Creates the ODBC data source connection

5. Runs the combined driver setup wizard silently

6. Launches and configures the monitoring application

7. Registers the ODBC connector in the Windows Data Source Administrator and configures the

corresponding DSN entries

The goal is a fully repeatable deployment that any technician can trigger without needing to understand

each individual step.

Tools: PowerShell, ODBC Data Source Administrator, Windows Registry, Microsoft SQL Server

Status: Partially functional. Troubleshooting in progress.