AI Systems Engineer | Security Γ Machine Learning Γ Intelligent Agents
π§ Email:
I build LLM-powered agent systems that connect reasoning, tools, and real-world data.
My work sits at the intersection of:
- π€ LLM Agent Frameworks (ReAct / Plan-and-Execute / Tool Calling)
- 𧩠Workflow Orchestration (LangGraph-style state machines)
- π AI for Security Systems (traffic analysis, adversarial robustness)
- π Graph-based Machine Learning for structured reasoning
I focus on building agents that actually run systems, not just chat.
Agent Frameworks
- LangGraph
- ReAct prompting paradigm
- Tool-calling / function-calling systems
- Multi-agent orchestration patterns
ML / AI
- PyTorch
- Graph Neural Networks (GNN)
- Contrastive Learning (SimCLR-style)
- Representation Learning
Systems
- Python / AsyncIO
- Docker / Linux
- Network data processing (pcap / flows)
- API integration & backend services
- ReAct-style reasoning loops (Think β Act β Observe)
- Tool-using agents (API / search / structured data retrieval)
- Multi-step planning + execution pipelines
- Memory-augmented agent workflows (short / long-term context design)
- State machine-based agent workflows
- Node-based task decomposition (Planner β Executor β Critic loops)
- Conditional routing between tools and reasoning branches
- Failure recovery + retry strategies in agent execution graphs
- MCP-style tool registration & dynamic tool calling
- Structured tool interfaces for external APIs
- Multi-tool coordination (search / analysis / reasoning / parsing)
- Safe execution control for agent tool use
- LangGraph-style DAG agent workflows
- ReAct-based iterative reasoning loops
- Tool routing + structured decision making
- Async execution of multi-step agent pipelines
- Graph construction from sequential / protocol-level data
- Hybrid systems: LLM reasoning + classical ML pipelines
β Thanks for visiting my profile!