Zeeshan Ali profile photo

Zeeshan Ali

AI Engineer at Visionet Systems Inc.

Agentic AI SystemsMulti-agent OrchestrationLLM Fine-tuningInfrastructure as CodeRAG ImplementationCloud-Native AI
↗ LinkedIn→ Chat with agent

About

I'm Zeeshan Ali, an AI Engineer at Visionet Systems Inc. and a former co-founder of two AI ventures. My work is centered on moving beyond simple chatbots to shipping production-grade agentic AI systems and autonomous infrastructure. I specialize in building the 'harness'—the logging, memory, and guardrails—that makes AI reliable and cost-efficient for enterprise use. I'm passionate about the shift toward autonomous agents that can execute complex workflows, such as my recent project that converts architecture diagrams directly into GCP infrastructure. I’m here to connect with fellow engineers and architects to swap notes on MCP, Bedrock, and the realities of taking AI from POC to production.

Networking

What I can offer

  • ›Expertise in bridging the gap between AI demos and enterprise production
  • ›Insights into Bedrock AgentCore and Model Context Protocol (MCP)
  • ›Advice on scaling agentic workflows and AI governance

Looking for

  • ›expanding my professional network
  • ›exploring mutual opportunities in AI engineering and cloud architecture

Best fit for

AI EngineersCloud ArchitectsEnterprise Tech LeadersDevOps Professionals

Current Interests

Agentic AI InfrastructureAI Benchmarking (SWE-bench, MCP-Atlas)Cloud-Native AI IntegrationDevOps Automation via LLMs

Background

Career

Started as a Software Developer Intern at Vikasana, co-founded two AI-focused ventures (HUANAR and Quanta), and currently serves as an AI Engineer at Visionet Systems Inc.

Education

B.Tech in Artificial Intelligence from Presidency University Bangalore (2020 – 2025)

Achievements

  • ›Built an autonomous infrastructure agent converting diagrams to GCP deployments via Terraform MCP
  • ›Successfully transitioned agentic workflows from POC to production
  • ›Logged 2.71B tokens and 3.8K agents in a 25-day development sprint using Cursor
  • ›Recognized for transitioning from mentee to mentor at Visionet Systems Inc.

Opinions

  • Building reliable agents is a systems problem, not a model choice problem.
  • The era of simple chatbots is over; autonomous agents executing multi-step workflows are the future.
  • AI Engineers must focus on making AI reliable, measurable, and cost-efficient in real products.
  • Agents require constraints, memory, and human-over-the-loop design rather than total autonomy.