Available for opportunities

AI/ML Engineer
Building Intelligent
Systems

Specialising in LLM integration, agentic AI, MCP server pipelines, and autonomous trading automation. Turning complex AI ideas into production-grade systems.

View Projects About Me
22%
Trade accuracy gain
15%
ROI uplift
99%+
System uptime
1+
Year production exp.
+91-6304986572 shashidharkarravula@gmail.com LinkedIn GitHub
Work

Featured Projects

🤖
AI-Powered Grid Trading Engine ⚡ Production
Multi-LLM decision making · self-healing architecture
⌄
CrewAIClaude AIRAG LangGraphMySQLTCP SocketPython
  • Built CrewAI agents (Trend Agent + Market Agent) to fetch live crypto prices and real-time market conditions, feeding structured context into a Claude AI-powered multi-LLM pipeline.
  • Engineered a Retrieval-Augmented Generation input system referencing historical grid parameters from MySQL DB, enabling LLMs to generate optimized Grid Bot inputs tuned to live conditions.
  • Designed TCP socket layer between LLM service and trade engine for low-latency delivery; self-healing loop triggers automatic LLM re-analysis on failure — achieving 99%+ uptime.
🛠️
AI-Driven Grid Trading Platform ⚡ Production
Multi-tool MCP architecture · live trade execution
⌄
MCP ToolsClaude DesktopRandom Forest RAGREST APIMySQL
  • Developed custom MCP tools (Market, DB, ML, Generation, Exchange) in Claude Desktop — users generate grid bot inputs and place live trades via natural language chat.
  • Built a self-improving ML Tool using Random Forest that retrains on every new grid input, combining live market data and RAG-style historical context for optimal parameter generation.
  • Engineered Exchange Tool with secure API/secret key management to execute live Buy/Sell grid orders from Claude Desktop — completing the full loop from chat to live trade.
📈
Asian Paints Stock Price Prediction 🎓 Internship
Time-series forecasting · historical data 2000–2021
⌄
LSTMARIMAGRU Neural NetworksPythonEDA
  • Developed a forecasting model using historical data from 2000 to 2021, achieving a 15% improvement in prediction accuracy.
  • Conducted exploratory data analysis (EDA) to identify trends and improve model performance across LSTM, ARIMA, and GRU architectures.
View on GitHub
🔍
Fake News Detection using NLP 🎓 Internship
Natural language processing · deep learning classification
⌄
NLP Deep Learning Logistic Regression PythonPreprocessing
  • Built a robust system for detecting fake news, achieving high classification accuracy through logistic regression and deep learning.
  • Implemented preprocessing techniques to optimize NLP workflows and ensure robust model performance across diverse datasets.
Career

Work Experience

AI/ML Engineer Apr 2025 – Feb 2026
🏢 Torus Kling Fintech · Hyderabad
  • Built MCP server pipelines integrating Claude AI for real-time crypto analysis and autonomous trade execution — cutting manual setup time by 80%.
  • Engineered EMA crossover, RSI/MACD multi-indicator signal fusion strategies — improving trade accuracy by 22% and ROI by 15%.
  • Developed self-healing event-driven agentic AI workflow using LangGraph that auto-recovered failed trading processes, achieving 99%+ system uptime.
  • Integrated REST APIs connecting data vendors, execution engines, and monitoring dashboards into a unified AI-powered trading platform.
Data Science Intern Nov 2023 – May 2024
🏢 AI Variant · Bangalore
  • Built NLP text classification pipeline and LSTM time-series forecasting model achieving 87% accuracy and 18% lower MAPE.
  • Developed Retrieval-Augmented Generation pipelines and agentic AI workflows; automated EDA reporting using Pandas, Matplotlib, and Power BI.
  • Delivered Asian Paints stock price prediction (15% accuracy boost) and a fake news detection system using deep learning.
Expertise

Technical Skills

AI / LLM
Claude AILangChainLangGraph CrewAIMCP ServersAgentic AI RAGPrompt Engineering
ML / Data Science
Scikit-learnTensorFlowPyTorch NLPLSTMARIMAGRU Time SeriesPandasNumPyMatplotlib
Languages & DB
PythonMySQL
Infrastructure
FastAPIREST APIsWebSocketTCP Socket
About

Who I Am

AI/ML Engineer with 1+ year of production experience building LLM-integrated systems, MCP server pipelines, and autonomous trading automation. Proficient in Python, LangChain, LangGraph, Retrieval-Augmented Generation, and agentic AI workflows. I love turning complex AI ideas into real, working systems that deliver measurable results — from a 22% gain in trade accuracy to 99%+ uptime on self-healing architectures.

🎓
B.Tech – Electrical & Electronics Engineering
Avanthi Institute of Engineering and Technology, Hyderabad · 2020 – 2023
Certifications
🏆
Data Visualisation: Empowering Business with Effective Insights
Tata Group · Forage
🏆
Data Science Certification
ExcelR