ali

jabbari pour

Project: Cloud Resource Management & Task Scheduling Engine Engineered a high-efficiency task scheduling system to optimize resource allocation in distributed environments. Key Implementation Features: Dynamic Mapping: Applied MCMF (Min-Cost Max-Flow) and Kahn’s Algorithm for task sequencing and workload balancing. System Resilience: Developed reactive logic for fault tolerance and real-time environment volatility management. SLA Compliance: Integrated EDF (Earliest Deadline First) heuristics to maximize hardware efficiency and meeting deadlines. Tech Stack: C# (.NET Core), Python, JSON Logic. Source Code: https://github.com/alijp84/AlgorithmProject

Project: MetaMind - Agentic Framework for Computational Intelligence Built an autonomous AI framework that uses LLMs to select and optimize 9+ Computational Intelligence algorithms (GA, PSO, ACO, etc.) for complex problem domains. Technical Achievement: Agentic Reasoning: Orchestrated a 7-step workflow using Llama 3.1 to analyze problems and auto-configure hyperparameters. Robust Engineering: Implemented structured JSON parsing from LLM outputs and a multi-run statistical validation protocol. Integration: Seamlessly connected Python back-end with OpenRouter API for real-time intelligent decision-making. Tech Stack: Python, Llama 3.1, Scikit-learn, OpenRouter API. Source Code: https://github.com/alijp84/Computational-intelligence-project

Project: 16-Bit Morris Mano Computer Implementation Developed a structural VHDL implementation of the Mano Basic Computer architecture, optimized for FPGA synthesis and modern hardware constraints. Key Technical Achievements: Advanced Bus Logic: Designed a Multiplexer-Based Common Bus to ensure FPGA compatibility and performance. Interrupt System: Integrated a Hardware Interrupt system with context saving and ISR vectoring. Control Unit: Designed a complex Finite State Machine (FSM) to manage hardware execution cycles. Tech Stack: VHDL, ModelSim, FPGA Architecture logic. Source Code: https://github.com/alijp84/CaProject

Cryptocurrency Trading System & Portfolio Optimization Developed a quantitative trading system in Python to backtest multiple strategies and optimize asset allocation using modern portfolio theory. Key Implementation Details: Algorithmic Strategies: Built and analyzed three distinct trading bots: Mean-Reversion (using EMA bands), Grid-Martingale, and ADX-based Trend Following. Backtesting Rigor: Implemented a rolling window validation (6-month training / 1-month trading) to ensure out-of-sample reliability and eliminate look-ahead bias. Portfolio Optimization: Applied Mean-Variance Optimization (MVO) to solve for Global Minimum Variance (GMV) and Max Sharpe portfolios across 5 crypto assets. Risk Management: Focused on capital preservation, achieving a significantly reduced max drawdown (-0.05%) compared to individual asset volatility. Tech Stack: Python, NumPy, Pandas, Scipy (Optimization), Matplotlib. Source Code: https://github.com/alijp84/trading_algorithm_project

DSL for Laboratory Data Analysis Developed a Domain-Specific Language (DSL) that simplifies laboratory data analysis by translating high-level commands into executable Python scripts. Key Implementation Details: Compiler Pipeline: Designed a 4-stage architecture: Regex-based pre-processing, AST generation (using Lark parser), Python code generation, and automated execution. Data Logic: Integrated automated data cleaning modules including IQR-based outlier removal and statistical filtering via Pandas. Visualization: Built a visualization layer to automate the creation of Histograms, Scatter plots, and Box plots using Matplotlib. Tech Stack: Python, Lark Parser, Pandas, Matplotlib. Source Code: https://github.com/alijp84/CompilerProject

Project: Milkit - Maternal and Infant Health Management System Developed a specialized software solution for systematic maternal nutrition management and infant health tracking, using a strictly typed architecture for clinical data insights. Technical Achievement: Architecture & Safety: Engineered a modular frontend using TypeScript (Strict Mode) to ensure architectural integrity and predictable state transitions. Centralized State: Designed a Centralized State Machine in the root component to manage the application lifecycle and data acquisition without routing overhead. Quantitative Nutrition: Built a computational engine to automate daily caloric requirement calculations and macronutrient distribution analysis based on user biometrics. Telemetry & Portability: Integrated a secure event-tracking system with ISO-standard timestamps and support for full system exports in JSON format. Tech Stack: React.js, TypeScript, Tailwind CSS, Vite. Source Code: Repository is private