Services / direct with Hamza
Freelance AI developer: add AI to your existing app
I’m Hamza El Haissouf, a freelance AI and full-stack developer working remotely with teams in any time zone. I add one useful AI feature to an existing SaaS, web or mobile app: the model is called from your server, the feature respects your user permissions, it is tested against agreed evaluation cases, and it has usage limits, fallbacks and human review before risky changes.
Key facts
- Who
- Hamza El Haissouf, freelance developer
- Where
- Casablanca, Morocco · remote worldwide
- Languages
- French, English, Arabic
- Price
- Quoted after a short call, based on scope
- Reply time
- Usually within a few hours on weekdays
- Last updated
« Un partenaire fiable et impliqué, que je recommande avec confiance. »
The problem
Most teams don’t need a new AI product. They need one journey in their current app to work better: summarising a record, drafting a reply, searching their own help content or classifying incoming requests. We pick that journey, write the test cases first, and integrate the feature behind your existing sign-in and data rules. Jobisque, CVFinder WinMatch and RocketApply show earlier model integrations; your feature is scoped and tested on its own terms.
01 / What this includes
What this includes
- 01
One AI user journey inside your existing SaaS, web or mobile app
- 02
Server-side model calls: API keys never ship in a browser or app bundle
- 03
Evaluation cases, usage and cost limits, fallbacks and review of risky writes
02 / How we work
How we work
Choose one journey and write evaluation cases
Integrate server-side with permissions and limits
Evaluate, add review and fallbacks, then release
Who this is for
- SaaS and app teams adding one AI feature to an existing product
- Founders who need a working integration rather than a demo
- Agencies seeking direct implementation support
Typical projects
- Record summary or reply draft inside an existing app
- Search over the product’s own help content
- Classification of incoming requests with human review
03 / A useful first scope
A useful first scope
Start with one user journey inside your existing app, using data the signed-in user may already see, with evaluation cases, usage limits and a fallback agreed before the build.
04 / What shapes the quote
What shapes the quote
The quote depends on the state of the codebase, integrations, data quality, the number of evaluation cases, review rules and model usage costs, which we estimate together from your expected volume.
05 / Relevant work
Relevant work
See the project record for the exact role, delivery state, and limits of this example.
MCP or a direct API? Read the comparison guide.
06 / FAQ
Questions before you start
Can you add AI to our existing app without rebuilding it?
Yes. I work in your existing repository, behind your sign-in, and add the feature as a testable module. We agree access, code review and acceptance criteria before starting.
Where do the API keys live?
Only on the server, in environment variables or your host’s secret manager. The web or mobile app calls your own API, which checks the user before it calls the model.
How do we know the AI feature works?
Before the build we write evaluation cases from real situations, each with an expected outcome. They are re-run whenever prompts, models or data change, so failures are visible before release.
How do you keep usage costs under control?
With per-user or per-account caps, timeouts, a model suited to the task and a usage view. Actual costs depend on the provider’s published pricing and your volume, so we estimate them together from your expected usage.
What happens when the model is wrong or unavailable?
The user keeps a non-AI path to finish the task. AI-proposed changes to customer records or outgoing messages are shown for approval instead of being applied directly.
Can you work with international clients remotely?
Yes. I work remotely with teams in Europe, North America, MENA and other time zones, using async updates, your Git repository and agreed review points. Calls and written updates are in English, French or Arabic.
Discuss your project
Tell me what you need. I'll suggest a useful first step.
Send a few lines about your goal, current tools, and timeline.
Project brief
Send a short project brief
Tell me what you need in a few lines. I'll reply with a useful next step.