UNDEFINED

Senior full-stack Developer working on high-scaling apps.

Freelance Software Developer

WYTLAND 🏝️

The next-gen Web3 community-owned travel platform that reshapes the outdated online distribution model built 20 years ago by a few monolithic & centralized platforms.

Tech stack :
• GitHub, CI/CD.
• Rest APIs microservices.
• Big data implementation.
• Docker services.
• AWS Infrastructure.

Chief Technology Officer

Jellysmack

Objectives and Key Results :
- Creating & staffing a brand new Computer Vision team.
- Product technical specification according to UX/UI guidelines.
- Technical architecture setup to automate Video editing / publishing for Creators.
- Improve Video editing process using Machine Learning and Artificial Intelligence.
- Clean code & Agile scrum day-to-day rituals setup.

Tech stack :
• Computer Vision (C++).
• Machine Learning (Python).
• GitHub, GitLab, CI/CD.
• Rest APIs microservices.
• Big data implementation.
• Docker services.
• AWS Infrastructure.

Senior Full Stack Engineer

Qwant

Objectives and Key Results :
- Creating & staffing the Qwant Music developers team.
- Data acquisition and modeling pipeline setup.
- Front & Back architecture setup.
- Clean code & AGILE Scrum rituals setup.
- Product technical specification according to UX/UI guidelines.
- Clean code & Agile scrum day-to-day rituals setup.

Tech stack : 
• API micro services back-end (NodeJS).
• WebGL API (C++ WASM).
• ReactJS front-end, Redux, Webpack, SSR, (S)CSS.
• Big data implementation (ElasticSearch) - Kibana - Grafana
• Git (hub & lab), continuous integration - deployment.
• Docker services.
• AGILE development methodology (using Atlassian Suite).
• User testing - UX.

Visit Live

• VR & Web / Mobile application (Samsung/Oculus).
• WebRTC implementations - C++.
• Web Sockets.
• Unity3D - C#
• WebGL API.
• 360° multimedia livrestreaming / rendering infrastructure (FFmpeg).
• Sysadmin (AWS cloud environment).
• R&D innovations.

Lead Developer

E-KIP MÉDITERRANÉE

• Project management.
• Mobile applications (Apache cordova, Angular/ionic).
• Web applications (symfony2, laravel, drupal, joomla, wordpress, php based, html5, css3/sass, javascript/jquery, python, bash).
• Network administration (cisco systems, nagios, centreon).
• Cloud computing (vmware esxi).
• Graphic design (about branding, advertising, visual identity, print).
• Security services (mainly proxies).
• Monitoring services - Java.

Software Developer

• Oracle SQL databases (Big Data).
• BiRT Reports - Predicting analysis with analytics.
• Linux & Windows server management.
• PHP intranet application (SAP connected).
• User training sessions.
• Helpdesk services.

Data Analyst Developer

🎒 Developer @ React • NestJS • Artificial intelligence 🤖 (OpenAI, LLMs, generative...)

I'm working with modern programming ecosystems on a daily basis using Typescript/Javascript to deliver cutting-edge solutions.

My proficiency in React & NextJS, and Nestjs, Fastify, Express allows me to develop versatile RESTful APIs and microservices to support seamless application functionality using CQRS & DDD principles.

I also developed a strong expertise in NodeJS,, which enables me to create sophisticated global enterprise-level web applications tailored to meet diverse client needs.

Depending on your goals, I can leverage low-level programming techniques with C++ and JavaScript to deliver high-performance Web Assembly applications that excel in speed and responsiveness across multiple devices. 🚀

I have an extensive experience in robust SQL and noSQL databases and Elasticsearch to handle large-scale data analytics and search operations. I emphasize comprehensive CD/CI processes (Docker/Kubernetes/Terraform) and rigorous quality assurance to ensure reliable and high-quality software delivery. 💼🔧

I like tackling both simple and complex challenges and aim to find the most elegant solutions possible 🧠. With an open mind and a thirst for ambitious projects, I am always eager to embrace new opportunities 🌟

Jean-Vincent Quilichini

Mimetix

We help your team to over-perform and eliminate all their repetitive processes.
Mimetix.co

Founder

The Incentive

We enable any Web3 project to reward users who tweet about them (Tweet2Earn). It's a win-win setup that leverages the power of communities for growth 🔥

[Sold]

Joyger

"At Joyger, we handle your welcome kits from A to Z!"
Many companies struggle with managing their welcome kits, and traditional vendors offer no solution to this problem.

Creating welcome kits is too time-consuming, from sourcing items to placing orders, packaging, and distributing them to each employee. That's why I founded Joyger to solve this issue.

A year and a half later, the company was sold.

HR Path

Developed an HR application using Flutter.
Managed and trained two new team members.

Lead Dev Flutter

CCA International

Customer Service Representative

McDonald's

Employee

At 14, I shattered the status quo by diving headfirst into entrepreneurship, and I've barely hit the brakes since 👊🏻

But it took me a while to figure out what impact I wanted to make on the world...

After a decade of launching numerous ventures, facing countless setbacks, and successfully selling two startups, it hit me :
Time is the most valuable asset we have ⏳

Somehow, we all end up wasting our time on tasks that just sucks.

My mission is straightforward: empower anyone to put an end to this never-ending cycle of tedious tasks thanks to replicative AI ⚡️

➡️ Streamline your processes
➡️ Eliminate repetitive tasks
➡️ Make your computer work on autopilot for you

Simply put, I want to make you more productive so you can gain back at least 2 hours each day—extra time you can use to enjoy life with your friends 🍹

Antoine Marcel

I didn’t talk about it on LinkedIn, but this summer I went to a Shaolin temple
1 month, 5 hours of training / day.

I got beaten with sticks like never before
And when I came back, like any good frustrated guy, I thought,
“Who can I beat up in return to get my revenge?”

But I’m more of a victim than a bully
And a proper geek
So I decided I’d beat up AIs until they overperform

Mission accomplished : last week we trained an STT model that’s 67% more accurate than Deepgram
Now people pay me to beat up STT models and make them better for voice AI on phone calls
(average +40% WER improvement and 300 ms latency).

Apparently, it helps startups close more deals,
move faster from PoC to production,
and cut their call error rates

So I’m happy, I’ll spend the next year taking out my rage the right way
✌🏻

building hyperspecialized stt that outperform any provider on the market ➡️ average +46% WER, 350ms

20 years ago, Sam Altman (AKA Samy) called me
He asked me how to make Whisper dominate the ASR market in 2025

My answer was clear
"We’re gonna eat you alive bro"
"Generalist STTs are garbage"
"With Latice(dot)ai we get an average of 45% more accuracy"

He’d been a bro for a long time
At least since preschool
So a bit shocked, he asked me

"How is that possible???"

But the answer is mathematical

"Specialized STT for one use case (vocabulary, call types, accents, bad microphones etc.) > Generalist STT"
"You can’t serve as well for meeting transcripts as you can for AI voice in hospitals"

I dropped the mic
And since then, I lost contact

After co-founding OpenAI, Mira Murati, the CTO, raised 2 billion at a 12 billion valuation
"Thinking Machine"

Her vision?
She believes specialized AI will overtake general AI.

Why keep chasing bigger models,
running on bigger GPUs,
with more data,
when your use case represents 0.00000...1% of that dataset?

Why use a model mostly trained on English when you’re working in Spanish?

She believes the future is specialization.
And so do I.

But for now Tinker is still on waitlist and only focused on LLMs :(
Too bad for all the AI voice founders struggling with STT.
I know that pain well: noisy environments, specific vocabulary, email address collections... All while keeping good latency. 
A nightmare.

That’s why we built Latice(dot)ai.
You tell us your use cases and where your current STT fails,
and we deliver a specialized STT just for you,
at least 30% better than any provider.

what I feel when a prospect tells me that 20% of their calls are a disaster because of their STT but then says their timeline is in 3 years 4 months and 3 days

Today I visited a trap house
you know, the kind of place where they sell drugs

I asked them, 
“How do you manage to serve 50,000 clients a day?
And what if suddenly you have 200,000…
don’t you lose money and make customers unhappy?”

The lookout brought me to the manager
The manager offered me some tea

very polite guy

and then he explained

They have an extremely well-optimized infrastructure
It’s all about auto-scaling

I can’t share all their secrets,
but what I can say is that now I know the secret to upscaling

Our STT infrastructure just went from handling 50,000 simultaneous requests
to an unlimited backend that scales in under 30 miliseconds
and can handle massive traffic spikes seamlessly

Thanks to the dealers
and see you soon

1 error on your voice AI STT can ruin an entire call
I’m sharing a case study that achieved +106% accuracy improvement on its STT

When your STT fails even once, the whole call can collapse
And it often messes up on critical details
-> Industry-specific vocabulary
-> Names, emails, addresses
-> Low-quality audio

But the end client doesn’t care about the tech
They just see 20% failed calls
And decide not to give you all their traffic
Or worse, not to renew the contract

For a medical-specialized STT, we released a full case study
-> WER by category
-> Cumulative WER distribution
-> Error type distribution
etc.

If you want me to send you the 9-page report
Comment “STT”

taplio

The Battle

Jean-Vincent Quilichini

Antoine Marcel

Followers

Posts

Engagements

Likes

Performance ScoreAntoine Marcel's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Jean-Vincent Quilichini

Antoine Marcel

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Antoine Marcel's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreAntoine Marcel's LinkedIn Performance