Data Analyst & AI Engineer
A data analyst and machine learning enthusiast from Indonesia, turning raw data into decisions that matter. I bridge the gap between analytical thinking and real-world impact — from predictive models to operational dashboards.
About Me
I'm a Physics graduate from Airlangga University based in Surabaya, with hands-on experience in data analysis, machine learning, and AI-driven solutions. My background in physics sharpened my analytical instincts — I love dissecting complex problems and distilling them into clear, actionable insights.
Currently at PT Berkah Tani Mukti, I build operational dashboards and data pipelines that drive real business outcomes — from identifying Rp 397.9M in cost savings to implementing customer segmentation models with 97% accuracy. Previously, I collaborated with PT Danone and Airlangga University on a real-time object detection system for workplace safety monitoring, achieving 90% accuracy and just 10 ms inference time.
I thrive at the intersection of data, technology, and business impact. Whether I'm wrangling messy datasets, training neural networks, or presenting findings to stakeholders, I bring both technical rigor and clear communication to every project. I'm always looking for opportunities to grow and contribute to meaningful, data-driven work.
Languages
Frameworks & Libraries
Tools
Databases & Platforms
Experience
PT Berkah Tani Mukti
Oct 2025 – Present
Built 8+ operational dashboards covering sales, logistics, fleet maintenance, returns, and receivables — uncovering Rp 397.9M in cost savings and exposing a 72% revenue concentration risk across key distributors. Analyzed field sales efficiency across Java regions, revealing a 3.12% YoY improvement and optimizing resource allocation for 10+ representatives managing 1,753+ retail partners. Implemented RFM segmentation and predictive models with 97% accuracy for 2,488 customers, reducing bad debt risk by flagging Rp 25.17B in critical receivables.
PT Danone × Universitas Airlangga
April 2025 – Sept 2025
Developed a real-time object detection model to identify ladder-holding individuals in workplace environments, achieving 90% accuracy. Implemented RTSP streaming for continuous safety monitoring and deployed the model using ONNX Runtime, achieving just 10 ms inference time for low-latency processing in safety-critical settings.
PT Semesta Integrasi Digital
February 2024 – June 2024
Generated 6 key insights addressing critical business problems and presented analytical findings to 4 stakeholders. Earned the Best Final Project title among 60+ participants. Developed an interactive e-commerce sales dashboard with dynamic filtering by date and region, enhancing sales trend analysis and boosting operational efficiency.
PT PLN Nusantara Power
February 2024 – May 2024
Conducted an in-depth efficiency analysis of PLTU power plants, processing 5,000+ rows of operational data. Identified 4 key KPIs to assess plant efficiency and detect operational bottlenecks. Evaluated system performance using Linear Regression, revealing optimal efficiency at a specific gross load from the High-Pressure Heater (HPH).
PT PLN (Persero)
January 2024 – February 2024
Developed a predictive model using SARIMA to estimate daily transformer load from 5 months of historical data. Identified load growth trends and potential capacity risks, ensuring compliance with the N-1 security standard. Achieved MAE of 0.4 for Trafo 1 and 2.4 for Trafo 2.
Bangkit Academy (Google, GoTo, Traveloka)
August 2023 – January 2024
Collaborated with Cloud Computing and Mobile Development teams to develop a color detection application for individuals with color blindness. Developed and optimized a machine learning model achieving 87% prediction accuracy. Converted the model to TensorFlow Lite for real-time mobile deployment. Graduated with an A grade (92/100).
Projects
Data Analyst · PT Berkah Tani Mukti
Designed and deployed an RFM-based customer segmentation system for 2,488 customers, paired with a predictive model reaching 97% accuracy. The system flags high-risk receivables — including Rp 25.17B in critical accounts — enabling the finance team to act proactively and reduce bad debt exposure significantly.
ML Engineer · PT Danone × Universitas Airlangga
Built a real-time computer vision system to detect ladder-holding individuals in industrial environments, achieving 90% accuracy. Integrated RTSP streaming for continuous monitoring and deployed with ONNX Runtime for 10 ms inference — making real-time safety compliance feasible at scale.
Data Analyst · PT Berkah Tani Mukti
Developed 8+ interconnected dashboards spanning sales performance, logistics, fleet maintenance, return management, and receivables tracking. The dashboards surfaced Rp 397.9M in actionable cost savings and revealed critical distributor concentration risks, directly informing executive-level decisions.
Machine Learning Specialist · Bangkit Academy
Collaborated across a cross-functional team (ML, Cloud, Mobile) to build a mobile accessibility app for color-blind users. Trained and optimized a color detection model to 87% accuracy, then converted it to TensorFlow Lite for seamless real-time inference on Android devices.