Al Roborol

Role

AI and Embedded Systems Engineer

Summary

I specialize in developing AI-driven solutions for embedded systems, with a strong focus on automation, performance optimization, and documentation. My work spans across system-level debugging, AI agent development, and collaborative initiatives that bridge real data with real value.

Key Projects

🧠 AI Initiatives and Embedded Systems

AI Agent for Documentation Automation: Developed an autonomous agent that interacts with Android devices via ADB to collect and update documentation.
AIxJIRA Projects: Led initiatives to summarize JIRA histories and predict anomalies using AI.
DevNexus Platform: Contributed to DevNexus.AI, integrating search, summary, and dialogue AI modules optimized for low-resource environments.

Skills

AI Development
Embedded Systems
Performance Analysis
Documentation Automation
Cross-functional Collaboration

Do LLMs Have Personalities? Or Even a Soul?

How Could LLMs Have Personalities? We know the personalities of our friends. But how to know an LLM’s personality, especially when they adapt their responses to match our input? (Perhaps Claude are somewhat sycophantic though.)

Why You Should Build AI Agents with Ollama First

The AI PoC Paradox: High Effort, Low ROI Building AI Proofs of Concept (PoCs) has become routine in many DX departments. With the rapid evolution of LLM models, more and more AI agents with new capabilities come every day. But Return on Investment (ROI) doesn’t change in the same way. Why is that?

Rethinking AI Agent Deployment: Start with Job Design, Not Just Technology

As companies accelerate their adoption of AI, many organizations are leveraging frameworks such as MCP (Model Control Platform) and A2A (Agent-to-Agent) from major LLM vendors. These tools support AI agent collaboration and contextual understanding, but are fundamentally LLM-centric rather than business-centric.

MCP: Standardizing How AI Interacts with the World—But Capability Negotiation Remains Unsolved

Even though MCP aims to be like a USB-C protocol, differences in knowledge and context windows between LLMs are fundamental barriers. Simply standardizing the exchange of function lists is not enough.