I build intelligence systems inside products and get them adopted inside enterprises.

My focus is on the stuff that determines whether AI becomes daily work or gets ignored: evaluation discipline, governance, workflow fit, rollout mechanics, and maintaining quality as things change.

Products shipped

🤖 Assist: AI operating system for marketing teams
Used by: Publicis, Toyota

🤖 BrandCheck: creative effectiveness and brand measurement
Used by: WPP, Google

🤖 PreFlight: AI video analysis and QC
Used by: WPP, Google

🤖 WAVE: AI-powered video automation
Used by: ThermoFisher

Product leader who sits close to the tech and closer to what users actually do.

Previously

🎬 Wooshii: founded to scale video operations and production
🧬 Envoke: co-founded, VR in life sciences
🍻 Hartest Brewing
🤪 Snowbombing Festival: co-founder

Patent: US 9754624 B2
Currently: CPO @ MSQ Global Studios
🏄🏽‍♂️ Surfer

If you’re turning AI into something teams trust and use daily, I’m always up for comparing notes.

Fergus Dyer-Smith

By day…
I am a senior data analysis professional with 15 years of experience in AI and Marketing. At LinkedIn I am helping our largest advertisers be more successful on LinkedIn using my deep AI and Generative AI expertise.
By night…
I explore, educate and share my passion around creative applications of generative AI for everyone with a fun and human touch with my AI Copilot Cat Teo.

By the numbers:
↳ 15 years of data analysis and AI experience
↳ 3 years of experience in Generative AI (images, video, sound, coding)
↳ 25+ workshops, 1000+ participants around AI and Generative AI at companies such as LinkedIn, Microsoft , Celonis and more
↳ 4 AI Workshops for Kids in schools, LinkedIn and more.

How can I help you?
↳ Workshops for adults, kids and companies in Generative AI especially around creative applications of GenAI (music, video, images, digital avatars) and Dangers of Deep Fakes
↳ Inspirational speaker ops around creative use cases of generative AI (Video) ad challenges (mis/disinformation)
↳ Have a question? Just Connect, Network, Learn Share, follow :)

Previous Speaker experiences around AI
↳ BITKOM (Germanys digital association): “Dangers of Deep Fakes”
↳ Zukunft Personal Keynote Panel : “AI and Work”
↳ 2x AI for Marketing Panel “AI and Marketing”
↳ Jury member for “Deepfake Prevention and Detection” Challenge (3.5M € over 1 year) for Germany’s Breakthrough Innovation Agency
↳ Various public speaking engagements around AI for LinkedIn Events

AI Workshop Highlights:
↳ 4x AI4Kids Workshops (Generative AI, Vibe Coding, AI for Education etc.)
↳ AI Workshops for Teachers (Poem, Song, Video, Image Generation Midjourney)
↳ Global AI Workshops at Microsoft, Celonis and more
↳ 25+ Workshops with over 1000+ participants

Awards Received:
↳ Named Top 100 AI Creative 2024 on LinkedIn
↳ Community Top voice for Artificial Intelligence in 2023
↳ Top 5 AI Innovators, Educators and Experimenters (12th July 2025) by the 5ives
↳ Global LinkedIn Sales kickoff Customer Science MVP for 2025 due to Contributions and Innovation around AI Applications

Get in touch? 
Ping me on LinkedIn. Happy to chat!

Stefan Kirschnick

"Don't serve ads, tell stories". My key message during our 2 Agency events in my Presentation around the B2B Video Revolution on LinkedIn. The other?

Beyond storytelling?

Meeting Face 2 Face in the Age of AI!
Still priceless. 
Lots of questions, discussions, ideas and "woah" moments, which I just love to share with wonderful people.

Fun annecdote:
In 4.5 years at LinkedIn. 
This was my first presentation in GERMAN on an event to multiple clients.
Even though we speak 99% english on the job.
We do often forget: its not our native tongue.
And... it felt awesome to express oneself (more?) naturally :)

Shoutout to Paul Banks and Nadine Jax for organizing 2 days for The Boutique Agency and The One Media GmbH and our clients :)

And to my co-presenters Leyla Yesilkaya and Julianne Veille <3

Again kudos to the wonderful piece of work around the Art and Science of (B2B)Video on Linkedin as well as the B2B Video Revolution by Ritika Gupta,Maja Rodic PhD,David Walsh,Wensy A., Jonathon Palmer and Bright Uduji

Have #FOMO and want to know the key 5 takeaways of the largest B2B Video Ads study?
Check out this 10 page PDF with relevant Insights around it:
https://lnkd.in/dJPsk-7e

Sr. Insights Analyst @LinkedIn | Generative AI Educator | AI Workshops + Speaker | Top 100 LinkedIn AI Creative 2024

LinkedIn #Hackweek Top 50 Finalist among 1100+ Projects globally! :)
Adrian +Stefan: "LiCollections - Supercharging Saved posts on Linkedin" ...

Think we have all been there... 
Found super cool posts on LinkedIn.
Saved for later.
100+ saved posts... can't really find it again... endless scroll... CTRL+F to search for it. Still can't find it. 

We are used to youtube Playlists, Twitter Bookmarks or Insta Collections...
Why not bring this to LinkedIn with a new LinkedIn twist?
Search, collections, tags and collaboration and discovery :)

How? 
Vibe Coding! Replit and more...
Bringing ideas to life with AI.

What worked?
+ pick an idea you love/is a personal paint point for you -> passion carries you forward
+ just put up an idea/project. Enables others to connect/discover and then you join forces 
+ Replit!!! Did surprise me very positively. Kind of crazy Agent mode vs. other tools. Great synch to github (even loading the codebase, not just saving it)
+ Hackweek Office Hours in DACH - a forum for folks to ask questions and ping pong idea (great idea Vincent!)

What didn't work?
- Had too many ideas. 
- Pinged lots of folks. 
- Then not having enough bandwidth and mental energy to follow through with most of them. 
- In German: "Asche auf mein Haupt".
- Better: either focus on one or make enough room in the week.

And... just from checking out all the different projects and finalists.
Magic happens when you connect people with ideas to tools that can bring them to life.
This is the promise of AI and Vibe Coding and the was the most exciting part to see :) 
SO MUCH CREATIVE ENERGY gets unlocked... beyond Engineering Folks!  

Big hug to Adrian Englhardt for the awesome work, being an amazing (hack)partner in crime and thinking big (beyond replit going into Claude Code and more ;)). 

Huge shoutout to the local DACH Hackweek crew, around our lead Vincent and partners in crime Pete Cholakov, Jacob Kaczmarek, Anja Willner!

(disclaimer: this is a POC/Hackweek. Does not mean that it comes to LinkedIn... but man I wish it did ;))

And P.S.: 
obviously had to give a couple of AI tools a shot to visualize it.
Reference Images our Hackweek Badge, Adrian and my LinkedIn profile pictures.

Best results:
+ Seedream 4.5 from Bytedance (most creative)
+ Google Nano Banana PRO (best text, prompt consistency)
Kinda meh:
+ Flux 2.0 and OpenAIs GPT1.5

#hackweek26

Anyone can relate? Countless 80% done AI Projects. Outdated after 2 months :). Well sharing one of them: Using Sora2 for a SKA Version of Bettina Wegners "Kinder" (1978)....

About the song:
Discovered that song and Bettina Wegner last year by chance.
Unfortunately the Lyrics are still very relevant today. 
And the song still deserves any ray of light it can get.

About the AI Video:
Speed of AI progress is... well pretty fast. So projects from 1-2 months ago quickly seem outdated. 
This one is me experimenting with Style References and Music Generation via Sora2 (from last year).

HowTo/Process:
+ create AI images in midjourney as visual style reference
+ upload the image to Sora2 
+ use my own cameo (basically my clone)
+ use a prompt to trigger style reference and to make Sora2 sing part of the lyrics.

Learnings:
+ good character consistency (myself in it)
+ Does a great job at synching instruments (drums/guitar). 
+ musically its not too bad 
+ Issue is music consistency as its only 10sec clips at the time with consistent singing and no audio upload

With Seedance 2.0 incoming.
This is probably gonna be even easier to do. (Audio Upload, Character References etc)

Prompt Used if you want to give this a try.
================
"Ein Splitscreen Music video.

 @STEFAN singt diesen song Text und spielt Trompete (rechter Splitscreen) und Tanzt Ska, während der dem Microphone Sing (linker Splitscreen). Music, Ska Punk Style. Beide schauen sich jeweils an. 

LYRICS:
[INSTRUCTION: This is the text that the LEFT part of the split screen sings, when it is over, he stops singing]
Sind so kleine Seelen
Offen und ganz frei
[INSTRUCTION: This is the text that the RIGHT part of the split screen sings, when it is over, he stops singing]
Darf man niemals quälen
Geh'n kaputt dabei

Zusatzkontext: 
+ Das Lied handelt davon, wie man Kinder behandeln und NICHT behandeln sollte.

Zusatzinstruktionen:
+ Finde für beide Teile des Verses (und somit beide Teile des videos) die passenden Szenen. 
+ Der ArtStyle des Videos soll dem ArtStyle des referenzbilders entsprechen!
+ Blende wo passend teile vom Text als captions/text overlays kreativ in das video ein.

Google keeps on shipping. Now Music generation via Lyria3 directly in Gemini.  30 sec clips. Multimodal input (image references etc.).
First tries?
Good quality... not yet Suno V5... but man... the way Google has been shipping recently , pretty sure they will get there soon.

AI Developer Experience at Google DeepMind 🔵 prev: Tech Lead at Hugging Face, AWS ML Hero 🤗 Sharing my own views and AI News

We just launched Lyria 3! Our most advanced AI music model in the Gemini app. 🎵

- Generates 30-second tracks from text or image prompts.
- Support custom lyrics, vocals, and cover art.
- Supports 8 languages including English, Japanese, and Korean.
- All outputs watermarked with SynthID for AI identification.
- Available for all users 18+ in English, German, Spanish, French, Hindi, Japanese, Korean and Portuguese, more to come

Try it now: [https://lnkd.in/dT2uurXf). Mobile will come over the next days.

Is Midjourney finally solving its biggest weakness? Text...
Here are some results from the unreleased Version 8 TEXT Rating Party... 

... biggest pet peeve I had with Midjourney up until now:
You just can't  make text work in images (almost) at all. 
Even simple words were a pain.

So it was hard to get the full creative power/style from Midjourney and make it useful in more complex images such as in ads without lots of postprocessing or using it together with Image generators that are stronger with text (so everything else) such as Nano Banana PRO.

Here are some interesting results from the Midjourney V8 TEXT(!) Rating Party, where  users are asked to rate pre-release images to tune the text capability of the model.

Important: 
Image quality is not the final image quality! 
(its supposedly significantly lower during the rating, so dont be offput by bad quality, focus on text)

From initial observations:
1. Significantly better than any previous Midjourney Version
2. Seem to be multilingual (German, Chinese etc. looked like they are working)
3. Seems to maintain creativity on text (engravings, graffity, logos, hidden text etc.)
4. Seems to follow more complex instructions for where text is put way better

Weaknesses so far:
1. lots of text (think Infographics) such as in Nano Banana PRO. Nope.
2. Not sure if the final model will maintain both creativity/style from MJ AND text coherence as well... This would be the dream.

And apparently we will soon know...

Their statement on when the release is:
"We'll be doing 3 more rating parties this week and then will be releasing V8 shortly after" (Status 16.02.2026)

Pretty clear what they are going for... Logos, Branding, creative engravings into images. 
More into commercial applications for ads etc. 

If you want to check out additional results/rate yourself head over to: 
(Current test ongoing: TEXT rating, not style, beauty etc.)
https://lnkd.in/dxyv3f6g

Looks promising overall!

taplio

The Battle

Fergus Dyer-Smith

Stefan Kirschnick

Followers

Posts

Engagements

Likes

Performance ScoreStefan Kirschnick's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Fergus Dyer-Smith

Stefan Kirschnick

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);}Stefan Kirschnick's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreStefan Kirschnick's LinkedIn Performance