Projects

A PostgreSQL-backed, append-only event store and enterprise audit infrastructure for multi-agent AI systems in regulated environments, where five AI agents collaborate on commercial loan decisions with full cryptographic auditability.

A LangGraph-based forensic multi-agent system that audits code repositories against technical rubrics using parallel detective and judge agents, producing objective evidence-based evaluation reports.

A credit risk assessment system that predicts loan default probability using machine learning, enabling data-driven lending decisions for financial institutions.

A time series analysis and forecasting system for Brent crude oil prices, combining statistical models with event-driven analysis to understand price dynamics and geopolitical impacts.

A multi-agent codebase intelligence system that ingests any repository and produces a living, queryable knowledge graph of its architecture, data flows, and semantic structure — designed to accelerate engineer onboarding in production environments.

A data-driven risk segmentation and premium optimization system for car insurance, using statistical analysis and predictive modeling to identify high-risk policyholders.

A Conditional Denoising Diffusion Probabilistic Model (DDPM) trained on MNIST to generate handwritten digit images conditioned on class labels.

An automated B2B outreach system that enriches prospects with real-time signals, classifies them against an ICP framework, and uses a LangGraph agent to handle replies, write CRM records, and book discovery calls.

Team PaLM's multi-database analytics agent built for the UC Berkeley DataAgentBench benchmark. Answers natural-language questions across PostgreSQL, MongoDB, SQLite, and DuckDB using MCP as the database access layer with self-correcting execution and traceable audit logs.

An agentic document processing pipeline that automatically classifies documents, selects the optimal extraction strategy, and executes structured data extraction using a multi-stage LangGraph agent architecture.

A data engineering pipeline that scrapes, cleans, and warehouses Ethiopian medical and pharmaceutical data from Telegram channels, enabling downstream analytics and object detection for medical product images.

A fine-tuned NER system for extracting product names, prices, and locations from Amharic e-commerce Telegram messages, enabling structured data extraction from unstructured Ethiopian marketplace content.

A quantitative finance system combining time series forecasting with portfolio optimization to support data-driven investment decisions and risk management.

An end-to-end machine learning pipeline for detecting fraudulent transactions using ensemble methods and advanced feature engineering on imbalanced financial datasets.

A MeTTa-based implementation of a classical rule-based system for clinical inference, demonstrating backward chaining to derive diagnoses from patient symptoms.

A cognitive-affective architecture port to the MeTTa language within OpenCog Hyperon, modeling internal drives, emotions, and goal-directed behavior using symbolic hypergraphs.

A computer vision system that extracts heart rate from facial video using remote photoplethysmography (rPPG), enabling contactless vital sign monitoring via a standard webcam.