Surpatanu acknowledges that current models are not yet reliable enough for fully autonomous agents, but argues that even increasing automation from 50% to 80% of a process would be a significant advancement. Tektonic uses a mix of foundation models and open models for entity extraction and lower-level actions. The startup is currently working with design partners to test and develop its system, with plans to become a SaaS company in the next three to five years.
Key takeaways:
- Tektonic, a Seattle-based startup, is emerging from stealth with a $10M seed funding round led by Point72 Ventures and Madrona Ventures. The company aims to combine generative AI with traditional symbolic methods to create workflow automation.
- The company's co-founder, Nic Surpatanu, believes that generative AI can bring a degree of adaptability and understanding of user intent to systems, something that was not possible with older tools like robotic process automation (RPA).
- Tektonic uses a combination of foundation models and open models for entity extraction and lower-level actions. The company is currently working with a number of design partners to test and build out its system.
- Surpatanu envisions Tektonic as a SaaS company in the future, connecting to APIs in customer systems. However, for now, getting started with Tektonic requires installing the system as a container in a business' virtual private cloud.