Roy Baladi | من تجربة قريبة من الموت إلى رحلة في عالم الذكاء الاصطناعي

في هذه الحلقة من بودكاست قرار، نستضيف Roy Baladi للحديث عن رحلة استثنائية تجمع بين التجربة الإنسانية العميقة والتحول المهني الكبير.

نبدأ من لحظة كانت قريبة جدًا من الموت، وكيف أثرت هذه التجربة على طريقة تفكيره ونظرته للحياة، وصولًا إلى بناء شركة تعمل في مجال الذكاء الاصطناعي والتوظيف في السعودية.

نتناول في هذا الحوار:

•  التجربة التي مر بها في لبنان وتأثيرها الشخصي 
•  كيف يمكن للحظات الصعبة أن تعيد تشكيل القرارات والمسارات 
•  بداية رحلته في مجال التوظيف والذكاء الاصطناعي 
•  مستقبل التوظيف ودور الذكاء الاصطناعي في هذا المجال 
•  التغيرات في سوق العمل في السعودية في ظل رؤية 2030 

هذه الحلقة مناسبة لرواد الأعمال، والمهتمين بالتكنولوجيا،
 وكذلك لكل العاملين في مجال الموارد البشرية والتوظيف الباحثين عن فهم أعمق لمستقبل هذا القطاع.

Transcript

SPEAKER_01 0:00

Ufajat and Shufit Infizure Rattakil Shang. Mam Bismah Helkalim et Mishma Ul Heshle Bedimar Yune Mabi Bahayat. Fashion Veto Hick, Mishma roof shoe, Ulesh, Ul Aisab, Ehlitalousir, canabit Ministry.

SPEAKER_02 0:23

When a kin had a fairian so, we share Jean Sau Farba Augustus, Ashirin Ashine, can fehs, Meshi Barabitu, Alabad Tulto Meat Mitribus, been Akbar in Fagar, Shefu Trich, El Hadis. Fakala Nefsu, Fus, Mordoakibir Gidden, who we said Arba Ashas Barabito Ashit Nel Arab Mustafa. Definitely has it halato in Loshachda has a lot of stories.

SPEAKER_01 19:24

A kid birloto. A kid birlato behara umasare, hasab strategy to eight is a B2B or B2C. B2B mahanda growth at any cost. B2C handak. My career from zero to twenty-six. So if an F Citigroup experience, keep a little under pressure. But let's see, clients, employees, and then we're not going to be able to do that.

SPEAKER_02 22:05

Oh smart recruiters in Baden. Anashufta and at what phase.

SPEAKER_01 22:30

Okay. You would connect to all the applications layers on metal job posting, job boarding, indeed, on the home base. No, I'm an accountant or accountant.com, echado amlo websites. So recruitment systems on premise. Ya name tale oracle, bitlaw CVs, that the status tabakel CV and Kabbal, Tash Shokhas, Makabal, Tashokhis, Muk Cloud. Centralized system. Exactly. Centralized system connect la job post job boards. LinkedIn indeed, career builder or monster. LinkedIn. Ba then you connect to assessment providers in video interview, behavioral assessment, skill assessment, onboarding, emission onboarding, reference check, background check. They're very weak with technology. Okay.

SPEAKER_02 27:35

And it forget beer for the market. And the show knows my Amelia digital transformation, operations that will fill industry.

SPEAKER_01 30:50

I just automated that piece. Travis Kalanikaba Uber said it best. Operations eats technology for breakfast. I don't know if he said it word for word. And then bottleneck, I'm top five people. Can you automate this? Automate your job. Find the auto operation process and automate your job. So Belish will outcome. Why didn't you prove how you automate everything down to how much money you need?

SPEAKER_02 31:48

Sensitivity, it's a good idea.

SPEAKER_01 31:56

Okay, VCs, startups, and we're going to be able to do that. Let's double down there. HR is HR. It's a cost center. You need to build something, you need to sell something. Any company. So build something is a revenue center, but that's your product. So investment had the brightest minds, product, so you don't have the highest level of tech in it. So the haikuna and you're going to have a very saturated market. The winners so far, the early applicant tracking systems, Taleo, Behat Bemiliar, Smart Recruiters, it's the next, hey, SAP, Central Modje. Okay. Big valuation. I think it's public mushaki. Probably not put it out. It's sold Central Modje S Plan Biahat. Okay. Smart Recruiters is the recruitment system of SAP. Now. So SAP, so minal winners, SAP Foj Elfino Asha, success factors, metal IBM, Oracle I forgot the name of it. But I think that's a good one. Bait, Bilblade Arab, Amlumya Sari. Not many more. And the trade paradox, and the trade skill assessments, eh? Matin Sesh was the Was of Kamehameha. It's still going. Eh Kamin. Was it is one of them.

SPEAKER_03 34:37

Yes.

SPEAKER_01 34:37

And the various tools, she 100 to 150 companies that exited grand max, in industry when thousands upon thousands upon thousands of companies. And so VCs are like the risk to reward ratio is low. Exits, not that many outside of smart recruiters, the paradox, in the last few years, not many. Dominated by LinkedIn. How are you going to be different than LinkedIn? How are you going to be different than indeed? So that's why. So what do we do as an HR tech entrepreneur? I think that you need to, yeah, now the future of work is completely changing. You look at Mesel and LinkedIn, you're going to get thousands of applicants. You're going to look at a database. It is still forces you to have a lot of humans to search. Individual messages, are you available? A debate dak Shura Tib. What's your nationality? Expected salaries, skill and test them. So right now all of this can be automated. If you can prove demonstrably, you know, meet Shahis, time to hire water, which is what what what Kenz is accumulating, hey, you can get investment. Oh, if you can prove and you have a different talent pool, you have AI enabled, AI-enabled talent pool, which you don't need meet Melieun Shaqis when you can have Khamsmid Elf AI engineers in a particular niche. Okay, it's valuable. Merkur is worth$10 billion. Kif amlo. They automated recruiting, focused on the large AI labs. They got excellent funding. Quality assurance testers, and they made a killing out of it.

SPEAKER_02 37:06

There is a lot of solutions that can come.

SPEAKER_01 38:58

What matters a daywait? Hell who was Shakzel Munesib. Time to hire. Time to hire, quality, and cost of hire, beautiful budget. I don't care how you solve it. You have a billion dollar company. Leno, Merkur,$10 billion company, bit lit this niin. Leno, they solved OpenAI, Elo, I need 10,000 data labelers. It's a marketplace. It's a two-sided marketplace. Wonderful. Then we know what product to build. So it's build the right thing and then build the thing right. Build the right thing has a true problem. So is it going to be a technology solution where AI my AI is likely a flight attendance? Well killed. Maybe flight attendance, tarif and no, and saudia, and fly a deal. But they tarify flight attendants, is the sort of thing. Where do you find them? Is it with competitors? Is it with flight schools? But then how do you qualify? What are they looking for? Is that trainers la flight attendants? Oh, you have one as an expert to understand what they have. Tatjon yamasal and training access events, digital events, tatwafer. So then you start to piece piece by piece. So this reverse engineering of the of the solution from the problem is a different way. This is product management. Any problem. Yeah, and it's so stu what how what problem are you trying to solve? Okay. What problem are you trying to solve? Start with why. What problem are you trying to solve? Why? Because flight attend because these flight attendants are leon. That's the why I'm doing this. Okay. The how I'm doing it by going to these areas where these flight attendants are needed and by ensuring that they have the skills that are needed. Why didn't you what? Had a whole product. Is that a belishable one? But the other thing is that the other thing is that the people who are not going to be tamiled. They're the ones who are going to do it. At some point, a CEO is going to be spread too thin. They need the CEO, she needs, he needs to be able to focus on what to do and how to motivate the team. Exactly. And that's an advice. So they were very generous. My goal is to make candidates happy. So general. So Anna Shameleth, Anna is a fresh grad. Fresh grad. Okay, so now I'm health marketing. Hab seen Alpha Dulge? Actor.

unknown 47:09

Okay.

SPEAKER_01 47:09

So I'm in Zaman, huh? Smart recruiters came in. Yane can best fit it, can a sum dollar in beer, but then times eleven. Yane for me will me. So Elli Messen make candidates happy. So Annabelle, our product release, but then nobody that shirke wake, make candidates happy, that will make you happy. Applicant, new applicant. So now I'll review stage. Interview. Is a Mahada Shaif? Is an Rafadat Belvishirki? Rahtarf, Laushuma Kane, the Shirki Maba of Fiat Malumet Annak. Can look an empty? Full transparency proumit private. Minimum. Garage, my garage, I don't care. I had nahed acted. Application. Put the resume send me a one click click apply at Dammit, and feed, share. So it doesn't matter if it will be garage, but then my Ulo, can feel any footageable garage though? It's a bad product. Jobs for Lebanon can it?

SPEAKER_02 58:26

So meaning the Oslan, and Osan Fakurf. Okay, Amalha, Hello. It's very interesting.

SPEAKER_01 59:02

It has to be sustainable. Ahamashi. So shall we have to be able to do that?

SPEAKER_02 59:19

Jobs for humanity, can it, Maamula?

SPEAKER_01 59:23

Jobs for human Tawa volunteer. Maximum security prison, Ladijal. Hagak Hagbira. Hagakbira, Akbad Hagat. Will and the gangs, no hope. I saw the Sud will be with Mexican Asians, leaders, Pelican Bay State Prison, Pelican Bay, al Katraz, and the other people who are not going to be able to do that. Okay. Okay, kiti Tamil gardening. Okay. Exactly. A moment of I want this. Exactly. Lel Amah. Dyslexia, ADHD, Dyspraxia, Tourette Syndrome, Dyskalculia. So Laelnafi no hatainabalami.

SPEAKER_03 1:04:18

Okay.

SPEAKER_01 1:04:19

Tadri Blackil Fa Min Hel Minhel Ashas no Soloma Sharikate, Wanatrib La Sherikate, mentorship, Min has Shericate Lah Albertina Alhamel, Shusl.com, a wallen, Menon, Virgin, the Keromano. So any homeless job seekers do can of lip denamian. Booking booking be tletin belladienisher and fikhi and solu mahalberti, programme, bitfa, bitfaw programme, wound sero shul a shot.

SPEAKER_02 1:04:55

Okay.

SPEAKER_01 1:04:55

Gamil Arabia, Bel Bid Larabi, Bakikens, U the Calestinae, but skill she talem to be jobs for humanity, breakens, yan is a hadan becans, ambition, can dar it mean a shudder, oh hadan leisure, Somalia, Tayro, ala Dubai, you can handle gap. L algorithm bin fehel she doesn't penalize the shakes, tests the skills. So, what is the thing? And can leisure but bonus is a cane leisure, Akian Obla Shikere, Wandon Corporate Social Responsibility, the same couldn't shaxel municip. Feed jobs for humanity and understanding of how to overcome barriers, will interview Lahadan doesn't want to make eye contact.

SPEAKER_02 1:09:41

Exactly. Firoya.

SPEAKER_01 1:10:01

And Baden You couldn't ballet Islam and Modern Islam. Tolana Tlital First Amal, Bel Bus Saudi, Haber Arbami Khamsin Shirke, Tsejelomana, Kilamelafiffetani Bellish and Ed Moose Lasu, Nistach the Met Deckal Sinait and Ed Nussel Oftal Shokisla Ofdalfursa. So the Messelin, Feriod, will budget about the Shamal Shakespeare. Zarif Arba Shinsack Tatajit Khalastibar.

SPEAKER_02 1:15:10

Gibb then Mosbut.

SPEAKER_01 1:15:54

Interview. Okay. Eh? Okay. Okay. For 15 minutes. Can it be showfuck? Will AI become fed? This is what's left. AI support human come in. Is it the top three? World class candidates. So he voice note e-commerce project manager. Triple checked, triple vetted, top three, my client, but detailed feedback. So transparency coming, only examples get criteria, uh, our top three criteria.

SPEAKER_02 1:17:20

A kid into the only criteria for you.

SPEAKER_01 1:17:35

Do you have the skills that are needed? And do you have it? You have the proper technical skills and experience? Do you have the soft skills? Are you someone who is a good thing? Would you have the soft skills? These are the three things. Have you done it? And you have the soft skills. What you do with this, you know, kill wadife and keys to success. So if you cognitive abilities, feminine research recruitment, but isn't it talent pipeline? And critical thinking. Oh, interpersonal skills is a negotiation of persuasion. So technical skills. So Analyze is AI One Assessment. Makes you more productive, Abu. Kibira gid.

SPEAKER_02 1:19:59

I think the podcast gives and see it humane. We'll uh by the way, you'll see I mean uh how we get this. Compliance compliance. Regulations, solutions, get then fatah, filament, uh capabilities without AI located, but regime tightnos fee and then obstacles.

SPEAKER_01 1:21:07

To have our own technology, we bayonet our data shared out on the PDPL data protection law, GDPR, so we in the bayonet, the Babi Saudi, the Bible server saw, and cybersecurity, uh, finance, healthcare, weekends, uh CVs, uh, mafia, server and garbage. So cyber security and compliance, soc one, soc two, and uh PDPL. Who made it server so our own large language models? And humane. Timing. It's enough. Open AI. Claude, X AI, MTL. Uh Gemini. Uh Claude. Claude was trained right after Open AI, National Gemini. Gemini wasn't being released out of safety. Open AI, U Claude, Nafel engineer, but it was trained on much better data. Data that was OpenAI, Belash Ken, Wikipedia articles, Reddit, also seeking the words. Gemini, alay, mate, Google. And they trained more and more and more on all types of data. This is Claude, Shusar, and Baidah, Malumet. However, on the training, inference, with large language models, vector transformer model, whose goal is to predict the next word, which means the coordinate, amus, kill kalimet laalam, will coordinates be known or ties with the model, sabamit plus coordinates. Hell coordinates, programming, language, that training phase is extremely important. Once you have the proper training, the right coordinates are you've mapped out the world, latitude, or longitude, best ala damon alkalimate, Clod, Akad, Malayin al Kutub, cover, Nasachun, Nasakh, Nasakon scan, scan, scan, scan, scan, scan, scan. The same as published books? Um I don't think so. Absolutely not.

SPEAKER_02 1:24:52

Mumkin Hagat Mishnah.

SPEAKER_01 1:25:06

Michael Jordan versus uh Kobe Bryant, who is the better player? Hammar Jordan is a Michael Jordan is a merrist, basketball, like Hammar. So coordinates, Hammar, Jordan, is it the Arab Allah Ambikod Sahif? You cannot train a model. Some have moved fast. Cloth. Clot code. So if you want to build something, you have two options. Either you build a foundational model, or I think we should. So we can win on this. But then the actual applications that come on top, and actor, actor, actor, actor, is stammel in stamelites. With a$5 million budget, they were able to train models on trained models. Is Amelia strategy? Because in terms of the application, we have to do that. To statement of cash flows, lahail business with tahit ma bad na shayatwa already. This is what I would recommend. Starting with that. Whether we have our large language model or not, they're going to build their next applications.

SPEAKER_02 1:29:19

So Mishnah, Mac Nordda. Thank you again. Thank you so much.

SPEAKER_00 1:29:40

Thank you, Ala. Shukran. That was so much fun. Therapeutic. Happy to hear. Yamil. Thanks, man.