AI Engineer with 1+ years in LLM/RAG Pipelines & Asynchronous Systems
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Results-driven AI Engineer with 1+ year of experience and a strong foundation in Mathematics, proven in architecting production-grade ML systems. Expert in deploying high-throughput LLM/RAG pipelines, API development (FastAPI), and orchestrating distributed asynchronous systems. Adept at transforming complex POCs into scalable, real-time architectures utilizing Docker, CI/CD (GitHub Actions), and GPU infrastructure. Fast learner with a track record of optimizing AI solutions for performance and cost, recognized as Best Presenter and MVP across top national AI bootcamps.
Diponegoro University
Bachelor of Mathematics · Mathematics
August 1, 2021 – April 1, 2025
NoLimit Indonesia
Data Scientist (Project-Based)
March 1, 2026 – Present
Bandung, West Java, Indonesia
Badan Pusat Statistik (BPS)
Data Analyst Intern
January 1, 2024 – February 1, 2024
Cirebon, West Java, Indonesia
MeetRecall AI – Intelligent Meeting Transcription & Analysis System
February 1, 2026 – June 1, 2026
Architected an end-to-end AI meeting intelligence system (MeetRecall AI) using FastAPI with asynchronous processing to handle multi-format audio/video ingestion, transcription, diarization, and analytics without blocking user interactions. Built an FFmpeg-based preprocessing pipeline to standardize MP4, MP3, WAV, and FLAC inputs into 16kHz mono WAV, ensuring model compatibility and stable transcription performance. Implemented a transcription and diarization pipeline combining Whisper-small and Pyannote 3.1 to generate structured, speaker-labeled transcripts in JSON format. Developed a semantic embedding and retrieval layer using paraphrase-multilingual-MiniLM-L12-v2 (384-dim) with FAISS indexing for efficient similarity search. Delivered advanced analytics features including RAG chatbot (FAISS + Qwen2-0.5B-Instruct), topic clustering (KMeans/HDBSCAN with LLM labeling), and speaker-level sentiment analysis (CardiffNLP RoBERTa).
View ProjectSmartSplit Bill AI – Intelligent Receipt Parsing & Bill Splitting System
January 1, 2026 – June 1, 2026
Built a web-based bill splitting application using Streamlit, enabling automated receipt extraction, participant assignment, and real-time payment calculation with total validation checks. Implemented AI-powered document understanding pipelines leveraging Donut (OCR-free Transformer), LayoutLMv3 (multimodal document model), and Gemini API, abstracted via a unified model loader for flexible experimentation. Designed structured data schemas and state management modules to convert unstructured receipt images into normalized transaction data (items, quantities, tax, subtotal, total). Containerized the system with Docker to ensure reproducible deployment and scalable architecture for future feature expansion (authentication, export, multi-currency).
View ProjectIntelligent Customer Assistant Chatbot
January 1, 2026 – June 1, 2026
Engineered a high-performance chatbot backend using FastAPI, enabling asynchronous request handling and scalable API endpoints for real-time customer interaction. Leveraged Groq's LPU inference engine to execute LLM queries with ultra-low latency, significantly reducing response time compared to standard CPU-based deployments. Orchestrated complex dialogue flows using LangChain, implementing memory management to maintain context across multi-turn conversations for a seamless user experience. Integrated a Retrieval-Augmented Generation (RAG) system with vector databases to fetch dynamic product knowledge, ensuring accurate, grounded, and hallucination-free answers. Designed a modular architecture capable of handling automated query classification, allowing the system to distinguish between general inquiries and specific support tickets.
View ProjectCNN Model For Plant Disease Detection PlantPal App
October 1, 2024 – December 1, 2024
Developed an AI-powered mobile application to identify plant diseases, provide tailored care plans, and offer expert tips. Built and trained a CNN model using Keras and TensorFlow, achieving 95.6% training accuracy and 97.2% validation accuracy. Deployed the model with TensorFlow.js for real-time plant disease diagnosis on mobile devices.
View ProjectHAKI Patent Modeling LSTM Deep Learning with GWO Algorithm
Unknown
May 1, 2025 – Present
Visual Quest Competition Finalist in Dataquest 3.0 Part of Airnology 3.0 from Airlangga University
Airlangga University
September 1, 2024 – Present
A silver medal at the Indonesian Mathematics Students Science Olympiad (ISSO 3)
Unknown
February 1, 2021 – Present
Cultural Fit Analysis
The candidate's project diversity, ranging from intelligent receipt parsing to meeting transcription and customer chatbots, indicates a broad interest in applying AI across different domains. Their involvement in academic competitions and bootcamps, along with awards, suggests a proactive and driven individual. The target role of 'AI Engineer' aligns well with their demonstrated skills in system architecture, deployment, and optimization of AI solutions. The project-based Data Scientist role at NoLimit Indonesia further reinforces their ability to work in dynamic, project-oriented environments.
Soft Skills & Operational Fit
The candidate demonstrates strong problem-solving, teamwork, and continuous learning abilities through project work and academic achievements. Their experience in orchestrating complex systems and collaborating with DevOps suggests good operational fit. However, the provided data does not offer direct insights into stress handling or team collaboration beyond project descriptions.