I turn ideas into
shipped products

Context

Datalent is a data education center staffed by experienced Turkish and Azerbaijani instructors from major global companies. The CEO wanted to expand the brand's reach, run awareness campaigns, and explore launching new academy branches in adjacent topic areas.

My Role

Executive Assistant to CEO (Feb 2024–Aug 2025). Despite the title, the role was operational: I led an intern team, owned project reporting and planning, executed brand awareness initiatives, and was given full ownership of the new branch launch process.

Process & Key Decisions

Over 6 months, I ran 3 brand awareness projects that engaged 250+ participants and generated 1,000+ leads, growing brand awareness by approximately 25%.

The main initiative was launching a new academy branch. We evaluated 3 directions: Prompt Engineering, Digital Marketing, and SMM. The first launch attempt (Prompt Engineering) failed — the team was mostly juniors and interns without enough execution experience, and it was summer, which killed lead generation.

Instead of abandoning the idea, I analyzed what went wrong, adjusted the timeline and team structure, and re-launched in a better season. Prompt Engineering won over the other two directions because it aligned naturally with Datalent's existing brand identity in AI and data science — it wasn't a stretch for the market to believe Datalent could teach this topic.

Outcome

The Prompt Engineering branch launched successfully on the second attempt. 20 students enrolled and graduated from the first cohort. The branch established a new revenue stream and validated Datalent's ability to expand beyond its core data science curriculum.

What I Learned

First launches fail often — what matters is the speed and quality of your iteration. Also: when choosing which product to launch, brand fit matters more than market size. Digital Marketing and SMM were bigger markets, but Datalent had no credibility there. Prompt Engineering was smaller but authentic to the brand.

Context

EESTEC is one of Europe's largest student engineering networks. The Baku chapter was relatively new and needed organizational reform, a stronger team, and recognition from the European body to achieve official Junior Local Committee status.

My Role

Chairperson. I owned strategy, team structure, partnership development, and the relationship with EESTEC's European board. I also served as Data Analyst for EESTEC Global, analyzing statistics across European branches.

Process & Key Decisions

The first priority was team reform — restructuring how members were recruited, onboarded, and retained. I focused on making projects the core value proposition: join EESTEC to work on real projects, not just attend events.

Key partnerships with SOCAR, STEAM Azerbaijan, and the Education Development Fund gave us credibility and resources that a student org normally wouldn't have access to. I organized the first BESTIEMEC joint event in Baku, connecting our chapter with the broader European network.

Outcome

Achieved Junior LC status (official European recognition). Grew to 1,000+ project applicants and 200+ active participants. Built a sustainable partnership pipeline that outlasts any single team member. Published promotional articles about EESTEC in Azerbaijani media.

What I Learned

Building an organization is product management at its core — your "product" is the community, your "users" are members, and your "metrics" are engagement and retention. The same PM principles apply: understand what people want, remove friction, and ship consistently.

Context

Azerbaijan has growing demand for Korean beauty products but no reliable local e-commerce option. Most buyers rely on informal Instagram sellers with inconsistent pricing and authenticity. Living in South Korea, I had direct access to suppliers and firsthand knowledge of the products.

My Role

Solo founder handling everything: market research, product selection, supplier relationships, Shopify store setup, theme customization (via Claude Code and Cursor), content generation, and go-to-market planning.

Process & Key Decisions

I built a three-stage Python pipeline: (1) hwahae_scraper to pull product data from Hwahae, Korea's largest beauty review platform, (2) hwahae_to_shopify to transform that data into Shopify's CSV import format, and (3) generate_content using the Claude API to auto-generate product descriptions in English and Azerbaijani. This let me go from 0 to 76+ import-ready products without manually writing a single listing.

The key decision was choosing automation over manual work. Writing product descriptions by hand for 76 SKUs would have taken weeks. The pipeline did it in hours and ensured consistency across all listings.

Outcome

Store built and approaching launch readiness with 76+ products, complete launch page copy, and a prioritized go-to-market checklist. Payment processing architecture designed (UK Ltd entity via Estonian e-residency for international payment gateway access).

What I Learned

Automating the boring stuff early is always worth it — even if the pipeline takes a day to build, it pays for itself immediately. Also: launching solo means you spend 70% of your time on things that aren't the product (legal, payments, logistics).

Context

Professor Brennan at Inha University was conducting research on tourism behavior — specifically how visitors experience Hanbok (traditional Korean clothing) rental at Gyeongbok Palace. The research involved multilingual interview data and social media content that needed to be collected and processed at scale.

My Role

Research Assistant. I was responsible for building the technical infrastructure for data collection and processing — not the research methodology itself.

Process & Key Decisions

Two deliverables: (1) A Whisper AI-based transcription pipeline that could handle multilingual interview recordings and produce clean, time-stamped text for analysis. (2) An Instagram scraper to collect public posts related to Hanbok/Gyeongbok Palace tourism experiences for content analysis.

The key technical decision was choosing Whisper over commercial transcription services — it handled Korean, English, and other languages without per-minute costs, and gave us full control over the output format for the research team's analysis workflow.

Outcome

Both tools delivered and used by the research team for ongoing data collection and analysis. The pipeline reduced manual transcription time significantly and enabled the team to process interviews in multiple languages consistently.

What I Learned

Academic research has different "product requirements" than startups — the priority is reproducibility and accuracy, not speed-to-market. Understanding the end user (researchers, not consumers) changed how I designed the output format entirely.

Context

Junction is one of Europe's leading hackathon organizations, but had never held an event in the South Caucasus. As Chairperson of EESTEC LC Baku, I proposed and led the bid to bring JunctionX to Azerbaijan — timed with COP29 to maximize relevance and sponsor interest around sustainability and AI.

My Role

Project Lead. I owned the entire operation: budget planning, sponsor acquisition, venue logistics, team recruitment, marketing, and coordination with Junction's global team in Finland. I also presented at Junction Asia's Regional Demo Day at POSTECH in South Korea.

Process & Key Decisions

The biggest challenge was credibility — we were a student-led LC with no track record of events at this scale. I built the pitch around COP29 timing and "AI in Green Economy" theme, which gave sponsors a reason to care beyond just a hackathon.

Key decision: partnering with the Education Development Fund and PwC Azerbaijan as anchor sponsors early, which de-risked the budget and attracted smaller sponsors. United Payment came on as a tech partner. SOCAR provided credibility in the energy/sustainability space.

I managed a cross-functional team across marketing, logistics, mentorship, and participant experience — all volunteers with varying commitment levels. The real PM work was keeping everyone aligned on scope and deadlines when nobody reports to you.

Outcome

350+ participants, successful 36-hour event, positive feedback from Junction global team. EESTEC LC Baku achieved Junior LC status (official European recognition) partly on the strength of this event. Presented the results at POSTECH in South Korea at Junction Asia's Regional Demo Day.

What I Learned

Leading without authority is the hardest and most valuable skill. When your team is all volunteers, you can't rely on hierarchy — you have to create momentum through clarity, energy, and making people feel ownership over their piece. Also: theme selection matters more than you think for fundraising.

Context

Fitness training in Azerbaijan is limited by high gym costs, few qualified trainers, rigid schedules, and no digital infrastructure connecting trainers to clients. People who want personalized coaching either can't afford it or can't find it.

My Role

Co-founder and Product Owner. I owned the product backlog, led user research, defined the MVP scope, managed sprint planning, and represented the product in incubation programs and pitch competitions.

Process & Key Decisions

We started with extensive discovery — 150+ interviews with gym-goers and fitness trainers to validate the problem. Key finding: trainers wanted better tools to manage clients remotely, and users wanted flexible scheduling above all else.

We entered five incubation programs (Sabah.incubation, Viveka Company Creation, IDDA Smart Solutions, TikTok Academy, Design Your Idea by Pasha Bank) — each forced us to sharpen the value proposition and business model. Being selected as one of 16 startups for TikTok Academy validated market interest.

The pivotal decision was choosing to focus on live 1-on-1 video sessions rather than a content/subscription model. Our research showed users valued the human connection with a real trainer over pre-recorded workouts.

Outcome

Reached MVP development stage with trainer partnerships secured. Graduated from multiple incubation programs with refined pitch and business model. Built a validated product concept backed by real user research data.

What I Learned

Talking to users before building anything is not optional — it's the entire job. Also, incubation programs are most valuable when you use them to kill bad ideas early, not just to collect logos for your pitch deck.