Code Story: Insights from Startup Tech Leaders S11 E27: Marco Rodrigues, Exaforce
Nov 18, 2025
Marco Rodrigues, co-founder and customer engineering lead at ExaForce, shares his journey from Canada to Silicon Valley and his passion for hockey. He discusses the motivation behind creating an AI-driven SOC platform to help overwhelmed security teams. Marco emphasizes the importance of semantic data models and collaborating with design partners to enhance the product roadmap. He also reflects on building a diverse global team and the future of AI-human analysts in cybersecurity, while offering wisdom for budding entrepreneurs.
AI Snips
Chapters
Books
Transcript
Episode notes
Agentic Security Combining Semantics And LLMs
- ExaForce fuses semantic understanding, statistical reasoning, and LLM decision logic into agentic 'exabots'.
- This combination enables analyst-level triage and response at machine speed with higher efficacy.
MVP Started With Data, Not UI
- Data acquisition, enrichment and pipeline design were the MVP focus for months before agent work.
- Without compressing and contextualizing logs, configs, and threat feeds, LLMs can't achieve high-quality machine-speed triage.
Behavioral Baselines Power Smarter Decisions
- ExaForce builds baseline behavioral models per identity and compresses relevant context into a 'knowledge model' per incident.
- That packaged context lets agents raise or lower alert priority accurately and quickly.
