Conference Presentation, Interview, Fireside Chat
Fireside Chat with Matan Grinberg, Founder & CEO of Factory | RAISE Summit 2026
- Factory released its current product version on September 26, citing a step-function improvement over previous iterations.
- Enterprise readiness is projected as the primary growth catalyst, with the expectation that organizations will be mentally prepared to adopt agentic AI for software development by the end of 2025.
- Matan Grinberg forecasts that open models will likely account for over 90% of future token usage due to superior speed, lower cost, and enhanced data control, while the share for frontier models is predicted to shrink from an expected 10%.
- Dynamic routing of engineering tasks to specific open models or frontier options like Gemini Flash is planned to optimize token usage management.
- The company's early market entry was identified as a strategic necessity to establish the team positioning required for current success.
- Future automation roadmaps include "droids" that monitor incoming communications (emails, Slack, texts) to identify priorities, auto-draft context-aware replies in the user's voice, and follow up on unanswered messages after three days.
- Additional "droid" capabilities are planned to ingest call transcripts to retain follow-up details and generate talking points for weekly all-hands meetings.
- Product expansion is expected to extend beyond software engineering into personal productivity and broader enterprise use cases, with existing customers like Adobe's finance team adopting the platform for non-development tasks.
- Users are advised to refrain from automating areas where business alpha resides and to be cautious of the diminishing returns of automating remaining low-leverage tasks after several months.
- Strategic planning involves implementing user limits once automation strategies reach critical mass to manage system load.
- Despite heavy reliance on automation, the speaker intends to maintain minimal technology usage during personal time, specifically reading paperback books without digital devices.
- Post-interview expectations include processing a volume of 100 emails, 100 Slack messages, and 100 texts to demonstrate the platform's capacity.