Hey, I'm Rama 👋
Data professional specializing in advanced SQL querying (CTEs, window functions, complex joins), Data Visualization using Metabase, and automated reconciliation pipelines using Google BigQuery. Focused on transactional, settlement, and statement data environments for B2B and fintech operations.
For Q&A, start a chat with
For escalations, please find my LinkedIn instead.
Tools I Use
GCP
Big Query
Metabase
Python
Javascript
Resume
Matchmade
Data Modeler
- Built an end-to-end automated reconciliation pipeline in Google BigQuery using advanced SQL (CTEs, window functions, complex joins), reducing manual reconciliation effort by ~60% for B2B clients.
- Reconciled transaction, settlement, and bank statement data across POS systems, payment switchers (Artajasa, Alto), payment gateways (Xendit), and corporate banks (BCA, BRI, BNI).
- Developed a reconciliation pipeline for retail clients (Pizza Hut), matching transaction-to-statement data to shorten reconciliation cycles from days to hours.
- Implemented SQL-based data validation logic for fintech clients (Bank Neo Commerce) to detect inconsistencies between core banking systems and payment switcher records.
- Optimized complex BigQuery queries to support high-volume daily transaction processing with consistent data accuracy.
Braincore
Machine Learning Engineer (Intern)
- Developed a computer vision liveness detection model, optimized and deployed using TensorFlow Lite (TFLite) for on-device mobile inference.
- Built a structured data preprocessing pipeline to clean and standardize large-scale image datasets, improving model training stability.
IPB University
Bachelor / Computer Science
- GPA: 3.19 / 4.00
- Relevant Coursework: Database Systems, Data Mining, Statistics and Data Analysis, Algorithm Analysis, Artificial Intelligence, Digital Image Processing.
Selected Work
Recent Projects
View all projects
Ram-AI Website Generator
An AI web app that turns text prompts into complete websites. Using Gemini LLM, I developed a tool that automatically generates HTML, CSS, and JavaScript files, making website creation incredibly fast and simple.
Viorama - Virtual Asistan
Developed Viorama - an intelligent library assistant system now actively serving UIN Sunan Kalijaga Library. This system leverages Gemini 2.5 Flash LLM with a multi-agent architecture to revolutionize how students and researchers interact with library services.
Rama's portofolio Facelift
This is my new portfolio featuring all the enhanced functionalities. What I'm most proud of is the comprehensive development work I've done across design, database architecture, and overall system implementation.
Plant Detection CNN
A machine learning model that identifies four plant types (corn, cotton, rice, wheat) with 90% accuracy. By using Python and TensorFlow, I designed a smart agricultural recognition system that can distinguish between different crops.
Personal
From the Journal
Open journalHello World!
Dec 14, 2025