Livia Ellen

- Building AI Automation engine and data pipeline for marketing team at NEOGOV, Imagine AI, omi AI, Syntervision - Optimizing model infrastructure for NEOGOV, PowerDMS, and NEOED (reducing AI pipeline generation cost by 60%)

now acquired by Element - Built ML models predicting vehicle transactions across 43K vehicles / 271 fleets - Forecasted fleet spend with time-series models; deployed API in Docker - Designed and deployed scalable Looker dashboards for real-time vehicle analytics

- Lead and manage data science and AI training programs for corporate clients and government agencies across APAC, the Middle East, Europe, Africa, and North America. - Impacted over 400 students globally, successfully launching 50+ projects. - Serve as the primary point of contact, ensuring compliance with program standards and providing ongoing support to foster student development. - Oversee multiple training batches in locations including Bali, Riyadh, and Kuala Lumpur, adapting to both onsite and hybrid models.

- Managed production data workflow in Dataiku in Pertamina PHKT Project: Well Portfolio Optimization which leads to 17% increase in ROI over the next 5 years for Attaka field – paper published in SPE Australia. - Developed R & Python scripts on Well Opportunity, Feasibility, Risk, and Analytical Engine, Enhanced data processing efficiency through a PDF Text Processing Workflow, reducing labor work: 2 weeks to 5 minutes. - Developed NLP workflows that reduced data processing time by 99%

• Deliver Data Management projects or services to clients • Supporting the Information Management and Data Management applications of Schlumberger Software Integrated Solutions (SIS) • Projects involved: - Data Scientist on Pertamina Hulu Kalimantan Timur - Well Portfolio Optimization - POV Data Science DELFI - Dataiku - DELFI Data Science Hackathon 2021 (developing http://id.slb-delfi-hackathon.com/ website and community chat for hackathon participant)

TL;DR : Started filterqu from a hobby building IG/Snap/TikTok filters-150+ branded AR effects later, we hit 10M+ organic UGC videos + 5B+ Impressions. Invited by Meta to launch their AR course, joined TikTok’s beta, and got early Snapchat Spectacles. Sunsetted after Meta deprecated Spark AR. Accomplishment I was proud of: - Develop interactive filter: AR Virtual Make Up, Quizz, Game on Facebook and Instagram using Spark AR studio, Lens Studio, 3D Blender & Javascript generating 4 Billion Impressions & 9 Million UGC, and significantly boosted client engagement by 412%. - Positioned as a key AR innovator, featured in 50+ international and national media outlets. Provided AR solutions for brands, enhancing their social media presence with interactive and engaging content. - Beta tester for Tiktok AR software - Effect House

- Designed and delivered the "Meta Spark Augmented Reality Pro Module" for Facebook Blueprint, highlighting the potential of AR and its future impact on the Meta Spark AR platform. - Guided learners through creative conceptualization, client brief interpretation, spatial design, and technical mastery of Spark AR features such as scripting, segmentation, and tracking. - Worked closely with a team to keep the curriculum current, relevant, and effective, ensuring a high-quality educational experience. - Engaged as a speaker and representative at hackathons organized by Facebook Developer Circles and Instagram, connecting with diverse tech communities across APAC.

Chosen as a leading global AR creator, I had the privilege to lead Meta's #MonthOfGood campaign internationally. This endeavor saw a collaboration with Raffi Ahmad & Rans Entertainment — prominent creators in APAC boasting a 60M-strong social media following. Together, we enhanced the spirit of Ramadan, offering users an immersive AR game experience. Post-campaign, the filter garnered a remarkable 5.5 million impressions, marking it as a significant success.

• Developing website, marketing promos & partnership using JavaScript, HTML, CSS in Hubspot: developing ruangguru main landing page, marketing related promos & partnership, ruangguru's product: BrainAcademy and ruangguru digital bootcamp. • Reducing human labor work from 14 to 2 days by developing AI auto-crop face function to process thousands pictures using python and matplotlib to detect peoples face position on pictures.

Relevant Coursework: Linear Regression, Computation for Data Science, Review of Probability & Statistics

In person hands-on course from Stanford combines classroom learning with guided visits to San Francisco’s leading fair and galleries led by Tatiana Maratchi Legrain understanding SF Art Investing

Bachelor Thesis: Student Engagement Detection System in E-Learning Environment using OpenCV and CNN • 3rd Runner Up Place Energy Innovation Challenge by Institution of Engineers Singapore (IES), • 3rd Place Astra Green Energy Innovation • 3rd Place Espriex Bussiness Model Canvas ASEAN • 1st Place Olimpiade Teknik UI – Idea & Research Proposal Competition • KEB Hana Bank Scholarship Excellence Awardee 2017