Selected work
Implemented systems,
explained honestly.
Each case describes the problem, my contribution, the design decisions, and the evidence available. Client identities are not disclosed.
The case studies.
- A bilingual AI sales system
- An anonymous case study of an English and Spanish AI sales system combining property context, conversation memory, qualification, and human handoffs.
- Coordinated browser automation
- How Younes Nadif designed a browser worker around task ownership, visible confirmation, and recovery when an action is uncertain.
- An operations platform with governed AI
- An implemented operations platform pilot connecting shared records, scoped AI tools, independent review, and version specific human approvals.
- An estimate journey connected to the team
- How Younes Nadif connected a service website’s initial estimate and later refinement to n8n and GoHighLevel contact and opportunity records.
- Document preparation with human review
- An anonymous document automation case study: extracting property information, mapping confirmed fields, preserving templates, and validating prepared outputs.
How to read them.
Every case separates what was implemented from what was measured. A pilot, a source review, or an inspected configuration is evidence of engineering; it is not a claim of revenue, hours saved, or current production health unless that was actually measured.
The walkthroughs are recreated explanations with fictional labels. They contain no client records, credentials, or private workflow exports.
The engineering guides explain the reasoning behind several of these designs in a form you can apply to your own tools.
What would a better
way of working look like?
Tell me about the task, the process, or the idea.
contact@younesnadif.com