Tatev Aslanyan

Senior Machine Learning Researcher

With a Master's degree in Econometrics, I bring 9 + years of robust international hands-on technical experience in Machine Learning, Data Science, and AI from Europe, US and Canada. High-impact global projects across Ecommerce, Recommender Systems, Search Engines, Marketing and more, with a focus on using Python, SQL, PySpark and cloud architectures like AWS. My skills extend to building and training Machine Learning and Deep Learning models in Python, complex SQL querying, cloud architecture, and econometric modelling. I am recognized for my analytical capabilities, statistical and mathematical background, and first-authored publications in top-tier US journals centered on NLP &amp; ML. An independent thinker. I thrive both autonomously and as a team player, complemented by my strong communication and collaboration abilities. Worked in Tech, as Data Scientist in ML and AI in both product-based and consultancy settings with a strong business sense, and collaborating with Tech Leads, UX designers, Senior Management, Product Managers, and Business Directors.🔍 Operate:Specialties: Machine Learning, Artificial Intelligence, NLP, Deep Learning, Advanced AnalyticsHave also Experience: Data Engineering, Product Data Science, LLMs(Bert)Innovation: Developed a recognized Data Science Bootcamp; Published first-authored papers in AI and NLP research at ACM🔗 Engage &amp; Explore:📘 Personal Portfolio: tatevaslanyan.com🚀 Data Science Bootcamp: lunartech.ai 💻 YouTube Channel: https://youtube.com/@LunarTech_ai?si=Xka8dNPM8Y2cB3S9📧 "The Data Science and AI Newsletter": https://tatevaslanyan.substack.com/💻 Github Code and Projects: https://github.com/TatevKaren✍️ Blog: medium.com/@tatevkarenaslanyan📆 Schedule a Call: tatevaslanyan.com/contact/Explore numerous Free Resources and Technical Code/Materials on my LinkedIn, Personal Website and GithubFeel free to explore the links above or drop me a message directly here on LinkedIn

Co-Founder

Co-Founder of LunarTech, and creator of LunarTechs "The Ultimate Data Science Bootcamp" ranked as one of the Best Data Science Bootcamps of 2023 [ITPro] featured on Forbes, TechTimes, and other international journals.

I created the following courses as part of LunarTech initiative:
🔘 Data Science Career Guide Course
🔘 Fundamentals to Statistics Course
🔘 Fundamentals to Machine Learning Course
🔘 A/B Testing Course
🔘 Introduction to NLP Course
🔘 Python for Data Science Course
🔘 Applied Data Science Course

No matter how many years of experience you have, if you don’t know how to treat people, it’s not “leadership.” It’s just a résumé. 🤝

- Skill gets you in the room.
- Character decides whether people stay in the room with you.

The older I get, the clearer it becomes: most “high performers” don’t fail because they aren’t smart. They fail because they create fear, ego, and silence around them.

If you want a simple leadership checklist:
- Speak with respect (even when you’re stressed)
- Give credit loudly, give feedback privately
- Keep your word
- Make people feel safe to tell you the truth and be free

Your title is temporary. Your reputation is permanent.

Don’t forget to follow LunarTech on social media for more AI & data science. 🤝

And if you want to switch careers (or level up fast): #LunarTechAcademy helps you become job-ready in AI + Data Science—with formal certifications in Data Science, AI Engineering, and AI for Executives.

Founder and CEO @ LunarTech | AI Engineer and Data Scientist | Seen on Forbes, Yahoo, Entrepreneur | Empowering Enterprises with Data Science and AI

There’s a popular idea floating around that early-stage founders should “activate their networks” to hire—former colleagues, friends of friends, people vouched for by angels or VCs. The argument is that using your personal network is faster, more efficient, and shows strength in “relationship capital.” I disagree—completely.

Early in our journey, I hired someone based on this exact mindset. A shared connection, someone from a familiar circle, vouched for their values and “good character.” That person turned out to be one of our worst hires.

No matter how well-meaning the referral was, the reality is:
▶️ Trust-based hiring leads to false confidence.
▶️ Familiarity becomes a substitute for skill.
▶️ And the cost of fixing that mistake is far higher—in time, in morale, and in culture—than simply hiring right through a proper, merit-based process.

I’ve seen this pattern repeat itself in startups, corporations, FAANG pipelines, and beyond. It breeds a dangerous environment:
• Some feel untouchable because they “know someone.”
• Others feel they weren’t hired on equal footing.
• Office politics starts to take root.
• And great talent—people who could genuinely contribute—is left on the shelf.

Culture gets damaged before it’s even built.

But it doesn’t stop at startups. This kind of network-based favoritism exists everywhere: in medicine, in justice systems, in public offices, and in every field where trust is misplaced and merit is ignored. You end up with unqualified people making decisions they have no right to make—while truly skilled, well-mannered, and capable individuals are unemployed or underused.

This isn’t just a hiring inefficiency—it’s a systemic failure. And we see the consequences daily.

So no, it’s not “faster” to hire from your network. It’s costlier. And it contributes to a long-term culture of incompetence.

If you’re a founder—even in the early days—build your brand. Let your company attract talent. Evaluate fairly. Hire through process, not connections. Trust your own judgment—not your friends’ opinions.

Merit > networks.
Process > shortcuts.
Earned roles > granted roles.

That’s the culture we’ve built—and it works.

This is my opinion, shaped by experience. And I stand by it! 

#LunarTech #culture #startups

No alternative text description for this image

We’re #hiring a Software Engineering Apprentice (AI Engineering Team) — Armenia 🇦🇲

This 6-month apprenticeship is for recent graduates and career switchers ready to build real production systems and AI-related projects.

Key details
- Remote-first for now (Armenia)
- 25 hours/week | Unpaid internship
- English B2 minimum (C1 preferred)
- Opportunity to be considered for a full-time offer after the program (not guaranteed)

What you’ll do
Build web + mobile apps, implement UI from UX designs, integrate - APIs/services, and help productionize AI modelsalongside AI engineers, UX designers, bioinformatics professionals, and data scientists across Western Europe, North America, Latin America, and Eastern Europe.

Tech (examples): Next.js/TypeScript, React Native or Flutter, FastAPI/NestJS, PostgreSQL/Supabase (pgvector), Docker, CI/CD.

👉 Apply via LinkedIn or send CV + GitHub/portfolio + short note to me! Comment interested below if the link is filled!

#aiengineering #hiring #lunartech

Have you ever been in this situation?

You’re sitting with a PDF or a handbook — an e-book for learning, a research paper, or a medical / legal document — and it’s in a language you don’t speak.

You need it in another language. Fast. And accurately.

You try Google and ChatGPT Translate.
You try copy–paste.
Everything breaks.

Formulas disappear.
Tables fall apart.
Images shift.
Footnotes get lost.
And the structure that actually matters is gone.

This is exactly why we built Babel.

Babel is a layout-faithful handbook + PDF translation engine.
It translates complex documents into your desired language while keeping everything else the same.

That means:
tables stay tables,
formulas (even complex math) stay formulas,
images remain in place,
headers, footers, references, and layout are preserved.

Only the language changes.

Yesterday, we open-sourced Babel and published our full test results.

We translated highly complex research papers (90–110 pages) across:
Arabic, Chinese, Spanish, French, Italian, Japanese, Dutch, Portuguese, Russian — and more.

What used to take weeks and thousands of dollars
now takes a few minutes.

We keep seeing the same use cases:
Students translating textbooks and handbooks for learning.
Researchers translating papers between Chinese ↔ English (and beyond).
Healthcare teams translating protocols and technical documentation.
Universities localizing academic material.
Anyone with an e-book who simply wants it in their own language.

If you have a research paper in Chinese and need it in English — Babel does it.
If you have an English handbook and want it in Armenian — Babel does it.

(For scanned documents, Babel Extreme handles translation + reconstruction.)

We open-sourced Babel because we believe language should never be a barrier to knowledge.

The tech behind it combines document-structure parsing, layout reconstruction, and AI-powered translation — built specifically for PDFs and handbooks, not generic text.

You’re free to:
use it, contribute, share it, and build with us.

Open source + tests on LunarTech Github

If you’ve ever struggled with PDFs across languages — this one’s for you.

About year ago I joined association. In my first presentation I mentioned my education and work experience outside🇦🇲 

A man objected. Then he asked something that’s actually worth answering:

“Why do we always hear about Armenian founders succeeding outside Armenia?”

So here’s my answer—based on watching, listening,quietly observing how “tech” works here

#Why Armenians succeed abroad

Because outside Armenia, there is usually higher chance that real value gets rewarded: competence, delivery, ethics, consistency, dignity. Not perfectly, not everywhere—but more often than in the current systems in Armenia

Inside Armenia, success can be filtered through a different logic: permission, not merit

let’s be clear: this is not about being extroverted. It’s not “go to more events.” It’s not “network harder.” What I’m describing is when ecosystem behaves like closed trust circle—where access depends on belonging, protection, loyalty chains more than on what you build. People are even afraid to say something because of the fear culture 

You see it in patterns people whisper about:
	•	same organizations & faces “rotating” through grants, sponsorships, visibility
	•	“community” platforms acting like gatekeepers, not enablers
	•	strong sexism & double standards still shaping who gets the spotlight and who gets “managed”

And if you stand up for yourself as a women and set boundaries- response isn’t “let’s fix it.”
It’s: “Let’s make YOU the problem.”
—while the “other founders” being showcased are already inside protective bubble. 

#The easiest example: AI-washing

Armenia loves to market “AI.” But you can literally feel the gap between branding and building when you meet the “AI” up close.

You’ll get a call from a so-called Armenian AI assistant, sounding futuristic… then you reply with your first sentence in English—
and suddenly the “AI” becomes a confused script that can’t handle reality.

That’s what happens when a system rewards labels, cheating, incompetence over competence.

#Davos - the uncomfortable reality check 🇦🇲

At Davos, Azerbaijan’s president Ilham #Aliyev said international relations are moving from the “rule of law” to “rule of strength”

If the world is becoming more brutal & competitive which is especially dangerous for Armenia, then in the very near future, fake companies, fake products, fake people, fake reality won’t save anyone

- Only real products, real progress, & real founders will

And that’s why this matters. The game being played right now in Armenia —where branding replaces building, - loyalty replaces merit—is a dangerous one in 2026 (some at the top who are involved know this - that’s why are perping their foreigner passports & exit plan) 

I’m not posting this to shame Armenia. I’m posting it because Armenia has real talent—and real builders including us deserve real ecosystem where value is the admission ticket, not “being trusted” by a cult.

LETS make real products, real connections & real impact! 🇪🇺🇦🇲

People love to blame “Gen Z” for poor work ethic or professionalism. I don’t buy it.

At LunarTech, we’ve worked with some outstanding Gen Z collaborators — curious, fast, reliable, and genuinely impactful. We’ve also seen unprofessional behavior from people across different ages, genders, and nationalities. In other words: it’s not generational. It’s individual.

A recent example: we opened a role for a grant writer. One applicant used most of the conversation to push “branding” instead (not what the role was for). Then came the pitch: “I’ll fix a couple of typos on your homepage for free… and you give me a testimonial.”

That’s not value creation. That’s a shortcut to borrow credibility — and to take credit for a brand founders build through years of daily work, consistency, and delivery.

The takeaway for founders and candidates:
	•	Respect the scope of the role
	•	Don’t waste time with misaligned pitches
	•	Don’t try to claim outsized credit for small cosmetic edits
	•	Be consistent: intentions show up in follow-through

Work ethic comes down to the person — how they were raised, their professional standards, and their intentions. Not their generation.

#Hiring #Startups #Professionalism #Integrity #WorkEthic #BrandBuilding

We’re #hiring: Grant Writer / Grant Strategist (Remote, Part-time)

LunarTech is looking for an experienced grant strategist to lead high-value funding efforts across AI / Deep Tech, with a focus on healthcare, education, and applied innovation.

- 100% commission-based (success-fee only) — no retainer
- Minimum 10 hours/week
- EU / US / Middle East funding experience is a plus
- Comfortable owning the full lifecycle: opportunity → strategy → compliant submission

If interested, DM me and I’ll share the role + survey.

Grant Writer / Grant Strategist (Remote, Part-Time, Commission-Based)

taplio