Paid campaigns measured to pipeline, optimised on lead quality, and governed by explicit budget rules.
Google and LinkedIn campaigns run with tracking readiness, honest reporting and lead quality analysis, not vanity metrics.
Google Ads
LinkedIn Ads
Meta Ads
TikTok Ads
View capability
AI systems and MCP
From a scattered process to one intelligent operating system
I design AI applications that understand a business process, call approved tools through the Model Context Protocol, move data between systems and return a traceable result. Claude or another model can work across HubSpot, analytics, internal databases, n8n and custom applications without turning the workflow into a black box.
01 / MAP
Model decisions, inputs, exceptions and permissions before writing automation.
02 / CONNECT
Expose only the right tools and data through governed MCP servers and APIs.
03 / ORCHESTRATE
Build multi-step workflows with validation, approvals, logs and recovery paths.
04 / IMPROVE
Measure quality, latency and business impact, then refine prompts and tools.
The result is machine-speed throughput and consistency on repetitive knowledge work, with human judgement retained where risk, ambiguity or customer impact demands it.
No mystery process. Six stages, each with a concrete output you can inspect.
01
Diagnose
Audit the website, tracking, CRM and campaigns against your goals. Interviews plus data, ending in a blunt findings list ranked by impact.
02
Design
Architecture before activity: what the system should look like, which parts to build or fix first, and what each phase must prove.
03
Implement
Hands-on build of the agreed work - website, CRM configuration, automation workflows, tracking - by the same person who planned it.
04
Integrate
Connect the parts so data flows once and works everywhere: forms into CRM, campaigns into analytics, automation across all of it.
05
Measure
Dashboards from first touch to pipeline with definitions everyone signs off. No vanity metrics, no unexplained numbers.
06
Improve
A standing cadence of experiments and fixes driven by the data the system now produces. This is where compounding starts.
The full delivery process, including what I need from you at each stage, is on theprocess page.
Service navigator
Tell me what you are trying to fix
You do not need to know whether you need "CRM implementation" or "marketing automation". The project brief asks about your goals, challenges and current tools in plain language, then maps them to the right engagement.
A step-by-step architecture guide for connecting Claude to CRM, analytics, databases and internal applications through secure Model Context Protocol tools.
The thought process and implementation method behind reliable AI systems, from choosing the right job through tools, evaluation, approvals, release and continuous improvement.
The practical 2026 shifts changing how growth and operations teams build AI, including MCP, governed tools, evaluation, focused applications and human approval.
Read insight ↗
FAQ
The questions buyers ask first
Can you manage both strategy and execution?
Yes, that is the point of working with me. From growth roadmaps and account-based campaign strategy down to hands-on work like building HubSpot workflows, developing websites or optimising Google Ads campaigns, I cover both. Strategy is written by the same person who implements it, so nothing gets lost in handover.
B
What industries do you specialise in?
I have worked across FinTech, MarTech, SaaS, education, logistics and travel, helping businesses scale their marketing and digital systems. The common thread is organisations that value data-driven growth and automation; the approach adapts to any sector that fits that description.
B
Do you work with international clients?
Yes. I am based in Tunis and work with clients across Tunisia, the Middle East, Europe and North America. Collaboration runs remotely through video calls, email and shared project tools, and I have delivered full engagements for companies in the United States and the Gulf without an on-site presence.
B
How do you measure success in your projects?
Clear KPIs agreed up front: traffic growth, qualified pipeline (MQL/SQL), acquisition cost reduction, ROI attribution and engagement, depending on the engagement. Dashboards and attribution models keep everything transparent, and I report against contribution rather than activity.
B
How does pricing work?
Three models: fixed-scope packages with published starting prices, monthly retainers for ongoing work, and custom engagements for larger programmes. Starting prices are on the pricing page; exact quotes follow a short discovery of your scope. No engagement starts without a written scope and price.
Describe your goals, your current stack and what is getting in the way. You will get an honest assessment of what I would build, in what order, and what it should achieve. No obligation, no generic proposal template.