Rahmat Fajri builds LLM systems that hold up in production.

Senior AI Developer

ESSA Industries Indonesia

AI Learning Engineer

PT Lippo Malls Indonesia

• Managed and optimized production-scale, on-premise serving of multiple open-source LLMs using vLLM for high-throughput, low-latency inference.
• Designed and deployed a multi-agent system for an AI Leasing Assistant (LangGraph), enabling 24/7 automated client responses and report generation for high-value leads.
• Engineered an AI-powered HVAC optimization system that combines real-time IoT sensor data with predictive ML models, delivering measurable energy savings and lower operational costs.
• Built an agentic RAG chatbot grounded in internal knowledge bases (SOPs, regulations) using LangGraph and Qdrant, significantly improving employee information retrieval accuracy.
• Architected a multi-tiered conversation memory system: Redis for short-term context and Qdrant for long-term retrieval, improving chat coherence across sessions.
• Applied an LLM-as-a-Judge evaluation framework to rigorously assess AI output quality and ensure regulatory compliance.

Data & Machine Learning Engineer

PT Pupuk Indonesia (Persero)

• Developed an AI-powered employee onboarding system using LangChain and Qdrant, reducing onboarding friction and improving access to information for new hires.
• Deployed open-source LLMs on Google Cloud Run with Ollama, balancing cost efficiency with scalability for internal use cases.
• Built and maintained CI/CD pipelines for automated deployment of multiple ML-serving APIs, cutting time-to-production for data science projects.
• Leveraged GCP services to deploy and maintain ML solutions with high availability and horizontal scalability.
• Designed a data ingestion pipeline with Prefect and Apache Kafka streaming data from multiple sources into Snowflake.
• Developed data models in Snowflake using dbt, enabling reliable analytics and self-serve reporting for business stakeholders.

Machine Learning Curriculum Developer

Dicoding Indonesia

• Designed and maintained a suite of data science and machine learning courses, improving curriculum quality and measurable student engagement.
• Applied NLP-based sentiment analysis to student feedback to surface actionable insights that directly informed curriculum improvements.