PROJECT BRIEF / AI AUDIT

Prepare a focused AI project brief.

Use this checklist before an AI audit, automation project, RAG build, LLM system or AI-assisted product sprint.

No physical office address is published. aicoding.am works as a service-area studio from Armenia.

Bring the process, data, users and constraints before choosing a model or writing code.

Intake checklist

Brief structure

What to prepare

Enough context for a useful first discussion

The business process, product goal or repeated operation you want to improve.
Current tools: CRM, ERP, spreadsheets, website, APIs, messengers, databases or document stores.
Users, roles, approval points and actions that must stay under human control.
Data examples, sensitive fields, access limits and deployment constraints.
What a useful first prototype should prove in one or two weeks.

Best first projects

Where AI engineering has a practical path

AI automation for repeated internal work, routing, enrichment or drafting.
RAG assistants for trusted documents, product knowledge or internal policies.
LLM systems with structured outputs, tool use and evaluation cases.
AI-assisted MVPs, dashboards, internal tools and integration surfaces.

What not to send

Risk and scope control

Passwords, tokens, private keys or production credentials.
Unredacted personal data before access boundaries are agreed.
A request to “add AI” without a process, user or decision point.
Claims about required model choice before the workflow is understood.