Machine Learning Engineer with 4+ years of experience building end-to-end AI solutions across Generative AI, Classical ML, and MLOps. Currently working at Insignia.co.id, delivering production-grade AI systems for enterprise clients in telecommunications, finance, cybersecurity, and pharmaceutical industries.
class Fahlevi:
role = "Machine Learning Engineer"
location = "Jakarta, Indonesia"
education = "B.Eng Engineering Physics - Telkom University"
focus_areas = [
"Generative AI (RAG, Agentic AI, LLMs)",
"Classical ML (Anomaly Detection, Feature Engineering)",
"MLOps & Production Deployment",
]
currently_exploring = ["LLM Evaluation (RAGAS)", "LangGraph", "Multi-Agent Systems"]🔹 Generative AI & LLMs — RAG systems, AI chatbots, multimodal document pipelines using GPT, Claude, Gemini, Nova, and open-source LLMs
🔹 Classical ML & Data Science — Anomaly detection, feature engineering, model benchmarking (LightGBM, XGBoost, BERT) on imbalanced datasets
🔹 Data Engineering — End-to-end ETL pipelines, big data ecosystems on AWS (Glue, Lambda, Athena, S3)
🔹 MLOps & Deployment — Automated retraining pipelines, model registry, batch inference on AWS SageMaker, real-time serving via Lambda + EventBridge
- 📈 Transitioned 2+ AI PoCs into production, generating IDR 580M+ in client revenue
- 🔍 Built Anomaly Detection system processing 42K+ security alerts with F1: 0.81 on 93/7 imbalanced data
- 📄 Developed AI-powered Document OCR achieving 99.17% metadata accuracy
- ⚡ Automated ML retraining pipeline running daily in ~13 minutes with auto model registration
- 🏅 Top 12 — Alibaba Cloud GenAI Hackathon 2025
- 🥈 2nd Winner — The Hack Deakin AI Hackathon 2021 (International)
Click to expand
Languages
ML & Deep Learning
LLM & GenAI
Cloud & MLOps
Databases
Web Frameworks
| Project | Description | Tech |
|---|---|---|
| Langchain-Ollama-JS | Integrating LangChain with Ollama using JavaScript | LangChain, Ollama, JS |
| Fine-Tune-Open-Source-LLM | Fine-tuning open-source Large Language Models | HuggingFace, Unsloth |
| Go ADK Agent | AI Agent built with Google ADK in Go | Go, ADK |
| Web Scraping API with AI | AI-powered web scraping API development | JavaScript, AI |
| CI/CD for ML Pipelines | CI/CD pipeline for ML models using GitHub Actions | Python, GitHub Actions |
| MLOps Zoomcamp | End-to-end MLOps practices and workflows | Python, MLFlow |
| Emotion Recognition | CNN-based facial emotion detection (35K+ images) | TensorFlow, Keras |
| LeetCode Practice | Algorithm & SQL problem solving | Python, SQL |
💬 Open to collaboration and new opportunities. Let's connect!