Applied AI Engineer · Open to international roles

I turn expensive manual work into AI systems.

The value is visible in the before and after: thousands of sales calls, hours of research reduced to minutes and complete content workflows run from one product.

View the proof Contact me
PRODUCTION CAMPAIGN3 phone lines active
CALLS COMPLETED3,00014 days
Collect
Enrich
Call
Qualify
Book
2 people identified the AI voice Human approval where it matters
Manual → 3,000sales calls completed in two weeks
Raw → 1,400accounts enriched, checked and scored
3h → <5mdaily market research workflow

More before-and-after proof

Different markets. The same economic logic.

Each case starts with a cost already paid in labour, lost leads or attention. The system moves that cost without removing the human judgment that protects the outcome.

01
Client operations

Loctician booking assistant

BEFORE

The owner handled every Instagram question, checked availability by hand and had to remember who needed a follow-up. Leads were lost even though demand was there. The pressure became serious enough to affect his health.

AFTER

The assistant answers in his tone, handles several conversations at once, books appointments and puts every lead in one dashboard.

MEASURED BENEFIT72 visits in 24 hoursThe Versailles rollout drew immediate traffic. Other locticians have already asked for the system.
SCALE CASEOne operator → concurrent conversationsThe booking engine can be deployed to other locticians with their own tone, calendar and rules. Demand no longer grows one DM at a time.
The owner can take on more demand without adding a person whose only job is to answer DMs and manage bookings.
02
Market intelligence

AI Dow Jones Analyst

BEFORE

A trader spent almost three hours collecting macro, positioning, volatility and media signals before each session.

AFTER

A scheduled agent assembles the evidence and delivers a readable Telegram briefing in under five minutes.

MEASURED BENEFIT3 hours → under 5 minutesAbout 60 hours of research returned to the trader every month.
SCALE CASE4 analyst workflows → 1 shared pipelineMedia, economic releases, market positioning and final synthesis already run through the same skeleton. A fund deployment would deepen the sources, controls and review layer.
Four U.S. financial analyst salaries represent $405,400/year in base pay before bonuses and overhead. As an analyst copilot, the target is 2–3× more research capacity.
03
Personal learning product

AI English Teacher

BEFORE

Three teacher sessions a week cost $400–500 a month, with feedback limited to what was noticed during the call.

AFTER

Five weekly voice sessions, mistake tracking and a study plan generated from the learner's own speech.

MEASURED BENEFIT95%+ lower monthly costFive sessions a week for about $20/month instead of three for $400–500.
$4,560–$5,760 saved per year compared with the teacher plan.

Flagship system · Outbound automation

From three disconnected sales jobs to one outbound system.

The system moves from a raw market to a qualified meeting without losing the lead between tools or teams.

01 / AI OUTBOUND SYSTEMProduction campaigns
PRODUCTION CAMPAIGN3 phone lines active
CALLS COMPLETED3,00014 days
Collect
Enrich
Call
Qualify
Book
2 people identified the AI voice Human approval where it matters
THE PROBLEM

More volume used to require more researchers, more callers and more appointment setters.

Lists arrived incomplete, research was repeated by hand, calls happened one by one and every booking created another manual update.

Lead Intelligence1,400 ranked accounts, 1,345 phone numbers and 1,246 owner names from one run.
AI Cold Calling3,000 calls in two weeks, live qualification and appointments booked automatically.
Loom GeneratorThree manual hours produced about 10 Looms. The pipeline can produce 300 personalised videos a day, with private pages and completion tracking.
CALL VOLUME3,000 callscompleted in two weeks
LOOM OUTPUT10 → 300/dayfrom three manual hours to parallel generation
VOICE QUALITY2 detectionsacross the full call run
20-REP BASE PAY$1.65m/yearU.S. benchmark before overhead
THE SCALE ECONOMICS

A human handles one call at a time. This architecture adds phone lines and workers without duplicating the research, qualification, booking or follow-up workflow. A 20-person service-sales operation represents $1.65m a year in U.S. base pay.

Model: 20 × $82,430, the May 2025 U.S. mean wage for a service sales representative.
BEFORE

Build lists → research accounts → call manually → qualify → book → record personalised follow-up.

AFTER

One system enriches, scores, calls, qualifies, books and generates personalised follow-up at campaign scale.

System family · Content operations

From scattered expertise to one content operating system.

The business benefit is the handoff removed between research, strategy, writing, production and publishing.

Long-form production

YTB Masterclass Factory

BEFORE

Research, scripting, voice-over, editing, thumbnails and publishing lived in separate tools and separate roles.

AFTER

One reviewable pipeline moves from a buyer problem to a finished, published YouTube masterclass.

BUSINESS BENEFIT0 → 10 qualified viewersThe first 6m 29s video was generated from script to final edit. Around ten viewers stayed for roughly two minutes on day one.
SCALE CASE$400–500 → automated productionThat is the quoted cost of motion design for one comparable video. The pipeline also handles the script, voice, thumbnail and upload. Its source method generated 200 qualified leads a month for its creator.
U.S. reference: a video editor earns $70,570/year median, before the other production work in this pipeline.
Knowledge engineering

Copywriter AI

BEFORE

Expert judgment was trapped inside 75 lessons. Every new sales asset still required hours of study or a senior copywriter.

AFTER

A reusable operating method now produces sales pages, VSLs and email sequences from the same decision rules.

BUSINESS BENEFIT75 lessons → 1 operating skill361k words became an 83k-word operating method that can be applied to every brief.
SCALE CASE$100k expertise → reusable skillThe expert behind the training charged $100k plus revenue share for a sales page.
U.S. reference: writers and authors earn $72,270/year median. The skill applies the same decision rules to every new brief.
Short-form intelligence

Content Creation Machine

BEFORE

Teams guessed what to publish, searched competitors by hand and started every script from a blank page.

AFTER

The system detects outliers, adapts the winning idea and produces the script, voice, captions and edit.

BUSINESS BENEFIT0 followers → 4,000 TikTok viewsThe same launch reached 1,000 Instagram views from a zero-follower account in a specific niche.
SCALE CASEOne format → parallel productionOnce the format is approved, the system can repeat research, writing, voice, captions and editing without rebuilding the workflow.
U.S. reference: writing and video editing represent $142,840/year in combined median pay.
Mohammed R.
MOHAMMED R.Applied AI EngineerBased in France · Available worldwide

About

The workflow comes first. AI carries the expensive repetition.

I find the work that burns time, salary or attention. Then I keep the judgment that matters and build the product around it so volume can grow without matching headcount.

Voice AILLM systemsData pipelinesRAGProduct engineeringEnglish · French

Open to international opportunities

Looking for an engineer who can own the full AI product?

shark.rhouta@gmail.com LinkedIn