AI DEVELOPER
PT Nawadata Solutions
March 2023 – Present
2+ Years
Designed and developed a scalable LLM chatbot backend using RAG architecture with LangChain, LangGraph, and FastAPI, supporting both local (Ollama, vLLM) and API-based models (OpenAI, Gemini).
Built a document parser and retriever API for a RAG system with ~95% retrieval accuracy.
Developed a near real-time fraud detection system using incremental learning and automatic scaling with Ray.
Created a flexible AutoML engine using FastAPI and Scikit-learn for various use cases.
Improved Anti-Money Laundering (AML) model accuracy by up to 90% through enhanced feature engineering and model selection.
Developed active and passive face liveness detection systems achieving sub-second recognition performance on millions of face images.
Built a prototype Applicant Tracking System (ATS) with spaCy and Sentence Transformers for NER and semantic matching.
Reduced manual data entry by 95% using NER and improved Vision OCR accuracy on damaged documents from 65% to 95%.
Collaborated with cross-functional teams to integrate prototypes into production systems, focusing on scalability and user experience.
Computer Vision Developer
PT Garis Posisi Nusantara
June 2021 – March 2023
1 Year 9 Months
Implemented Mask R-CNN on drone and satellite imagery for building and vehicle detection as part of a Geospatial Intelligence project.
Developed automated PDF reporting tools with instance segmentation maps and heatmaps, reducing processing time by 90%.
Implemented U-Net for land use classification and automated PDF reports with segmented maps and area analysis.
Created a system to extract ship positions from Maxar satellite imagery for regulatory monitoring and analysis.
Led a team to deliver a GIS-based Geospatial Intelligence project on time and according to client requirements.
Trained over 20 client staff on GIS tools, improving their technical capabilities.
Worked with defense and security stakeholders to gather requirements and deliver customized GIS solutions.
Ensured code quality and maintained project documentation for long-term scalability and collaboration.
Speaking Engagements
Guest Lecturer
Universitas Catur Insan Cendikia
Machine Learning and LLM Course
February 8, 2025
Delivered an in-depth lecture on Large Language Models (LLMs), covering their architecture, fine-tuning, and deployment.
Shared real-world development practices and technical insights from hands-on AI development experience.
Provided practical guidance on building production-grade AI systems.
Public Speaker
Universitas Catur Insan Cendikia
HIMATIF Anniversary Event “Artificial Intelligence in Shaping the Future”
December 21, 2024
Presented strategies for thriving in the AI era, including career preparation, mindful AI usage, and avoiding AI dependency.
Engaged the audience with a dynamic Q&A session focused on AI opportunities and challenges.
Delivered key insights to help students and professionals navigate the AI-driven future.
Educations – Skills – Language
Telecommunication Engineer
Institute Technology Telkom (Now Telkom University)
Final Project: License Plate Number Recognition Based on Template Matching and Cosine Similarity (Matlab)
2015 – 2019 | GPA: 3.74/4.00