blog
Articles on AI security, large language models, RAG architectures, and Web3 engineering.
-
Your Open-Source AI Model Might Be a Sleeper Agent
I trained four backdoored coding models and evaluated probe detectability across latent activations to catch sleeper agents. Two of three defenses worked—here is what happened.
-
Building dApp on Ethereum Blockchain using Flutter and React
Architecting decentralized applications using Solidity smart contracts, Flutter cross-platform mobile frontend, and React Web3 integration on Ethereum.
-
LLM Finetuning: Enabling Quality Embedding and Text Generation for Amharic Language
Parameter-efficient fine-tuning (PEFT/LoRA) and tokenization strategies to enable high-quality Amharic text generation and semantic embeddings.
-
Precision RAG: Prompt Tuning For Building Enterprise Grade RAG Systems
Engineering production-grade enterprise RAG systems with Weaviate vector database, prompt tuning, semantic reranking, and hallucination reduction.
-
The End-to-End Web3 dApps: Certificate Generation, Distribution, and Value Transfer with Algorand
End-to-end decentralized credentialing, NFT certificate minting, and verifiable value transfer built on the Algorand blockchain.
-
Redash Chatbot Add-on: LLM-based Chatbot for Advanced Data Analytics and Visualization
An LLM-powered natural language interface for Redash BI, enabling automated SQL generation, query visualization, and conversational data analytics.
-
Data Warehousing: The Engine of Business Intelligence
Designing scalable ELT data pipelines, star schema architectures, and automated Apache Airflow DAGs for business intelligence and traffic monitoring.
