Suggestions
Raphael Derbier
Director of Product, Principal Solutions Architect, Innovation leader
Raphael Derbier is a seasoned professional with a passion for leveraging complex technologies to drive productivity.
With a background as an innovation program director, product manager, and cloud architect, Raphael has led distributed teams across Europe and America.
His expertise in enterprise software architecture, software product development, and project management has honed his technical proficiency, customer-centric approach, and leadership skills.
Eager to embrace new challenges, Raphael is seeking opportunities where his blend of creativity, technical prowess, methodological expertise, leadership acumen, and people-centric focus can make a tangible impact, ideally in roles such as product manager, solution architect, or innovation team lead.
Key to Raphael's success in his career are his insatiable curiosity, unwavering passion, steadfast commitment, and relentless pursuit of knowledge.
Raphael Derbier pursued his education at Télécom Paris and has amassed a wealth of experience through roles at reputable organizations across various capacities.
His professional journey includes stints as Staff Architect at Hypermode, Director of Product at Dgraph Labs, Senior Principal Solutions Architect at TIBCO Software Inc., Presales Strategic Solution Group at TIBCO Software Inc., EMEA Technical Presales Leader & worldwide BPM expert at TIBCO Software Inc., Managing Director at Solucom, Managing Director at Vistali, Solution Architect at WebMethods, Consultant at ALTEN, and Developer at NMS - Natural Micro System.
Highlights
Product recommendation using RAG on Dgraph. Before going on vacation and forget everything, I crafted an educational content using my beloved graph database Dgraph, small language model for semantic embeddings and LLM for entity extraction and text generation.
The post has a Jupyter Notebook companion so you can experiment all the steps yourself.
Happy coding with Dgraph, embeddings, and language models!
https://lnkd.in/gsdGeDjm
“A.I, It’s like magic”. I recently heard this comment about the output of a generative A.I. I certainly experienced this feeling myself but I’m not sure it triggered the same thoughts. Just sharing mine here.
Some people continue to quote what is called the Clarke’s third law, from science fiction writer Arthur C. Clarke: ”Any sufficiently advanced technology is indistinguishable from magic.”. I don’t like it for the main reason that I like both tech and magic”. Magic produces an emotion, an astonishment, a moment of poetry, triggering sometimes a smile. This is because something is happening that does not fit with reality, with our common sense, something surprising. The disappearance of a bird in the opening scene of the movie “The Prestige” is perhaps my favorite example to illustrate the gap between the effect and the means used to produce this astonishment: the magician makes a bird disappear from a cage and reappear somewhere else. We later understand that the bird has, in fact, disappeared, crushed by the heavy mechanism of the folding cage. The one we think is reappearing is just another bird that looks like the same. The trick is nice, it hides a horrible secret requiring the sacrifice of a bird each time it is performed.
With magic, confronted with a well executed performance, I usually enjoy being fooled and stop there.
It is very different with technology. Magicians do not share their secrets. Engineers and researchers have usually the obligation to share theirs with their peers for review and validation. As many parts of my life are influenced by technology, I find it important to be able to understand how it works. If something is ‘technologic’ I should be able to find someone who can explain it to me, if that’s not the case,I have a doubt about the dirty secrets that might be involved.
What would be your preferred role facing AI ‘magic’?
- The audience: enjoying the show with no interest in knowing the details of the trick.
- The illusionist: finding smart (and sometimes not avowable) ways to surprise the audience.
- The engineer: searching acceptable solutions and explaining details in a way that people can understand so that they can decide what to do about it.
I’d like to be in the third category. It’s not easy!
