Akshansh Agrawal@kokushibho
2
Events attended
2
Submissions made
India
1 year of experience
About me
Akshansh is a student with a strong interest in technology, programming, and innovative project development. He enjoys exploring areas such as Data Structures, Operating Systems, Artificial Intelligence, Machine Learning, and software development. He has worked on and explored projects including Deadlock Detection and Recovery, Deepfake Detection, CPU Scheduling, Satellite Collision Analysis, and AI-based systems. He focuses on understanding how technologies work practically rather than only learning their theory. He is particularly interested in building projects that combine programming, algorithms, and real-world applications. His goal is to continuously develop his technical skills and create useful, practical software solutions.
🤝 Top Collaborators
.png&w=640&q=75)
.png&w=640&q=75)
.png&w=640&q=75)
🤓 Latest Submissions
SENTINEL : Autonomous Options Trading Agent
The system consists of five primary agents. The Research Agent continuously scans the market and analyzes stock prices, trading volume, implied volatility, historical volatility, options volume, open interest, liquidity, and market trends to identify potential opportunities. The Strategy Agent evaluates these opportunities and selects an appropriate options strategy, such as bull put spreads, bear call spreads, iron condors, or other defined-risk strategies. It calculates important metrics including maximum profit, maximum loss, probability of profit, expected return, and risk-reward ratio. Before any trade can be executed, the Risk Officer Agent evaluates the proposed position against predefined risk controls. These controls include maximum risk per trade, portfolio exposure, daily loss limits, position size, liquidity requirements, bid-ask spread limits, and overall portfolio risk. The Risk Officer has the authority to reject any trade that violates the configured risk rules. Once a trade is approved, the Execution Agent communicates with Alpaca through the MCP server or CLI and submits the order to the dedicated paper trading account. The Portfolio Agent then continuously monitors open positions, tracking profit and loss, market conditions, risk exposure, and exit conditions. When a predefined profit target, stop-loss, or risk condition is reached, the system can automatically close or adjust the position. SENTINEL also maintains a detailed record of trading decisions, agent reasoning, executed orders, portfolio performance, and trade outcomes. A real-time dashboard displays portfolio value, P&L, active agents, market regime, open positions, risk status, and the latest AI decisions. The complete workflow is Research → Strategy → Risk Validation → Execution → Portfolio Monitoring → Exit.
4 Sep 2026
.png&w=828&q=75)
AI anomaly detector
The README.md file is the main documentation of the AI Memory Leak & Resource Anomaly Detection Platform. It gives developers and users a complete understanding of the project, including its purpose, features, architecture, technologies, installation process, and future development. It starts with an Overview, explaining that the application monitors system and process resources, detects abnormal memory behavior using Machine Learning, and uses a native C backend for performance. It then describes the major features, such as live CPU and memory monitoring, Isolation Forest-based anomaly detection, the Tkinter dashboard, real-time graphs, CSV logging, and automatic Python fallback. The Project Architecture section explains how the major components communicate. main.py starts the application, the logger handles application logs, backend.py manages the native C backend, monitoring modules collect telemetry, the anomaly detection module analyzes the data, and app.py launches the Tkinter dashboard. The README also documents the Technology Stack, including Python, C, Tkinter, Matplotlib, NumPy, Scikit-learn, and ctypes. The Folder Structure shows where important files such as main.py, backend.py, monitor.py, anomaly.py, utils.py, and the C source files are located. The Installation and Running sections provide the commands required to clone the repository, create a virtual environment, install dependencies, and start the application using python main.py. It also explains how to compile the native C backend for Windows, Linux, and macOS. The Machine Learning Pipeline explains how raw telemetry moves through feature extraction and Isolation Forest anomaly detection before producing an anomaly score and possible memory leak detection. The README also describes logging, generated reports, the development roadmap, contribution process, license, authors, acknowledgements, and future scope.
10 Aug 2026
