Identifikasi Walet
A computer-vision system that distinguishes individual swiftlets in noisy CCTV footage through detection, embeddings, and identity classification.
Loading interactive visual
Thariq Ivan Anendar
AI / ML engineer in Surabaya, Indonesia
AI systems that workbeyond the notebook.
From CCTV frames to power-plant time series, I turn models, data, and software into systems that can be measured, shipped, and used.
28 secindustrial pipeline / from 1+ hour
99.33%bird ID / test accuracy
0.99997individual ID / ROC AUC
0.01–0.05 cmcalibrated measurement / error
Systems in practice
Built, measured, and put to work.
Each case study traces the engineering decisions behind a working result, from computer vision and retrieval to industrial data and autonomous systems.
See the complete project archiveLawBot
A document-grounded legal assistant for Indonesian TNI law revisions, combining fine-tuned language models with hybrid retrieval and reranking.
Generative AI Saham
A market-forecasting application that connects five years of price history with news sentiment and LSTM time-series modeling.
Engineering principles
The model is only the middle of the story.
I’m drawn to problems where research judgment and implementation discipline matter equally. My work spans machine-learning models, vision pipelines, retrieval systems, industrial data, embedded systems, and web products.
The model is one part of the job. I carry the work through evaluation, integration, deployment, and the interface where someone can actually use it.
- Based in
- Surabaya, Indonesia
- Education
- Institut Teknologi Sepuluh Nopember
- Degree
- Bachelor of Informatics Engineering, 2022–2026
Start with the decision
Every experiment should resolve a technical choice or improve an operational outcome.
Measure what matters
Accuracy matters alongside runtime, error bounds, reliability, and the change users can feel.
Finish the system
A useful model belongs in a dependable pipeline, a clear interface, and a realistic deployment path.
Work under real constraints
Where engineering met operations.
Industrial data, UAV development, and technical problem-setting—three environments where reliability mattered beyond a demo.
Methods behind the outcomes
Depth across the AI stack.
Model development, retrieval, vision, data, product engineering, and embedded systems—connected by the work they enabled.
From model selection and fine-tuning to evaluation that reveals whether an approach is ready for a real system.
- PyTorch
- TensorFlow
- Keras
- Scikit-learn
- Transformers
- QLoRA
- CNN
- LSTM
- ResNet
- ConvNeXt
- InsightFace
Work recognized beyond the repository
Trusted, funded, and tested in competition.
A record spanning national UAV competition, funded research, and technical distinction.
2025
Bayusuta · TEKNOFEST Turkey
Funding Recipient and Team Member
Autonomous UAV system for real-time forest-fire detection and mapping.
Recognition 1 of 6: Funding Recipient and Team Member, Bayusuta · TEKNOFEST Turkey, 2025.
Learning that compounds
Recent technical study
- June 2026
Microsoft Azure for AI and Machine LearningMicrosoft
1KLL40LG1J9M - June 2026
Foundations of AI and Machine LearningMicrosoft
GEVTXJMCPF1X - February 2026
Databases and SQL for Data Science with PythonIBM
WTIH3TJIT034 - 2024
Deep Learning: Neural Network & AIUdemy
Independent course - 2023
Natural Language Processing: NLP With Transformers in PythonUdemy
Independent course
Open to ambitious technical work
Bring me the hard problem.
I’m interested in AI, computer vision, data, software engineering, and research roles where technical depth must become a dependable system.