https://sulthanabiyyu.github.io/portfolio/ Recent content in Portfolio on Sulthan Abiyyu Hugo -- gohugo.io en-us Sat, 20 Jul 2024 17:06:01 +0700 https://sulthanabiyyu.github.io/portfolio/anak-baik/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/anak-baik/ An Indonesian dataset for teaching language models to tell ethical instructions from unethical ones. Published at SEALP 2025. https://sulthanabiyyu.github.io/portfolio/marzuki-7b/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/marzuki-7b/ Continued pretraining, SFT, and DPO alignment on top of Mistral-7B. Three checkpoints published. https://sulthanabiyyu.github.io/portfolio/oneanalytics/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/oneanalytics/ A single place for the whole company’s business metrics, replacing a request-and-wait reporting process. https://sulthanabiyyu.github.io/portfolio/ai-image-classification/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/ai-image-classification/ Detecting whether an image was generated by AI, using fine-tuning techniques on ResNet. Published in JTIIK. https://sulthanabiyyu.github.io/portfolio/herman/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/herman/ An Indonesian adaptation of the Hermes function-calling dataset, for training models to emit reliable structured output. https://sulthanabiyyu.github.io/portfolio/analytics-chatbot/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/analytics-chatbot/ An LLM that answers business questions directly against company data, grounded hard enough to be trusted. https://sulthanabiyyu.github.io/portfolio/artinya/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/artinya/ Translation that tries to preserve tone, nuance, and cultural meaning, then evaluates whether it actually did. https://sulthanabiyyu.github.io/portfolio/dustiness-qc/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/dustiness-qc/ Video analysis that measures the dust plume from falling powder, turning a subjective QC judgment into a number. https://sulthanabiyyu.github.io/portfolio/rec2slide/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/rec2slide/ Reconstructs a slide deck as a PDF from a screen recording. Published on PyPI. https://sulthanabiyyu.github.io/portfolio/receipt-invoice-ocr/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/receipt-invoice-ocr/ Automated extraction from receipts and invoices, built to run on edge devices. https://sulthanabiyyu.github.io/portfolio/etl-pipeline/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/etl-pipeline/ A dockerized ETL pipeline with star-schema dimensional modeling, built to compare processing engines. https://sulthanabiyyu.github.io/portfolio/pipeline-reliability-qcc/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/pipeline-reliability-qcc/ Took a critical pipeline from recurring SLA breaches to full compliance, and won the group-level QCC. https://sulthanabiyyu.github.io/portfolio/kuantil/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/kuantil/ Option pricing and quant primitives as small, tested Python modules. https://sulthanabiyyu.github.io/portfolio/fanesis/ Mon, 01 Jan 0001 00:00:00 +0000 https://sulthanabiyyu.github.io/portfolio/fanesis/ Maps relationships inside a group from images, using face recognition and emotion detection. https://sulthanabiyyu.github.io/portfolio/data-engineering/ Sat, 20 Jul 2024 17:06:01 +0700 https://sulthanabiyyu.github.io/portfolio/data-engineering/ Since 8 months ago, I began learning everything I could find about Data Engineering. I’ve consumed a lot of information on the topic from books, blogs, videos, courses, and forums. I’ve learned about the fundamentals, data warehousing, and data movement techniques like ETL and ELT. It has been quite a challenge, but also a lot of fun! There’s still so much more I want to explore. Data Engineering involves a multitude of tools, and I’m curious about every one of them.