Sign In

Scott Persinger

CEO at Supercog AI, prev at Tatari, Stripe, Heroku, founder at CloudConnect

Scott Persinger is the CEO and co-founder of Supercog AI, a startup focused on solving Application Integration using LLM (Large Language Model) powered agents.13 Supercog AI was launched in public beta in 2024, with the goal of applying the GenAI stack to problems in the modern workplace.1

Professional Background

Prior to founding Supercog AI, Scott Persinger held several notable positions in the tech industry:

  1. Vice President of Engineering at Tatari, a data and analytics platform for TV advertising.2
  2. Leader at Stripe, where he built the alternative payment methods engineering team.2
  3. Senior Engineering Director at Heroku, leading the platform services engineering group.2
  4. Co-founder and CTO at Cloud Connect, which was acquired by Heroku.2
  5. Co-founder and VP of Engineering at Remixation, acquired by Vevo.2

Education and Expertise

Scott Persinger graduated from the University of California at Berkeley with a Bachelor's degree in Computer Science.2 His expertise lies in engineering leadership, startups, and application integration.35

Current Focus

At Supercog AI, Persinger and his co-founder Alex Osborne are working on connecting LLMs to various systems used in workplaces, aiming to unleash the power of AI reasoning engines to help accomplish real tasks.1 Their goal is to create a platform that solves tangible problems by integrating AI with both structured and unstructured data in business environments.1

Scott Persinger is active on various platforms, including GitHub, where he maintains 66 repositories4, and his personal blog, where he shares insights on startups and engineering leadership.5

Highlights

Aug 7 · linkedin

This is fun. I did a quick session with Claude this morning to create a bash script which sends "git log" output to GPT4o for summarization. I had to go a few rounds with Claude to get the script working, but now I have a nice bash script which will generate a very readable summary of a git commit, without me relying on the author having to write it.

Grab the finished script here: https://lnkd.in/g2uY_Q2D

Here's an example of the output:

% bash ~/Downloads/git-log-summary-script.sh -9 Summary of git commit (index -9):

Commit Summary

Author: ALEX OSBORNE [email protected]   Date: Fri Aug 2 22:23:19 2024 -0700

Description:   This commit focuses on enhancing the functionality for detecting Markdown tables in the codebase related to the EditorState class. The table detection logic has been made more robust and includes new functionality to identify the start position of a Markdown table. Additionally, references to OpenAI usage statistics and authentication configurations have been removed from the AdminTool but are still included in the list of available tools.

Key Changes:

  1. Markdown Table Detection:   - Enhanced the detect_markdown_table static method to not only detect the presence of tables but also to track their starting position in the Answer object.   - Updated the method to handle line-by-line checking for Markdown table headers and delimiters.   

  2. Code Cleanup:   - Removed configurations related to OpenAI API keys from the AdminTool class, improving security and simplifying the handling of sensitive information.

  3. Increased Max Agent Time:   - Increased max_agent_time from 180 to 500 in the AgentState class, potentially allowing agents to run for a longer duration. ... #genai #practical-ai

Aug 5 · linkedin

Are we in the middle of an AI arms race and we just don’t know it? This is retired U.S. General Paul Nakasone, who was appointed to the board of OpenAI in June.

This announcement came quickly on the heels of the news of the departure of Ilya Sutskever and the entire safety team at OpenAI. A member of that team published the “Situational Awareness” essay, which attempted to characterize the massive investment which is going into AI and make some startling predictions about the likelihood of us achieving AGI by 2027.

One could certainly see the appointment of Nakasone as tacit confirmation of the national security implications for current AI research.

I have had a long fascination with the Manhattan project and the early development of nuclear arms. One of the keys to that history was that while the science underlying nuclear technology became quickly understood around the world in the 1930s, the creation of nuclear weapons required a huge industrial process.

It is important to see the historical parallels in the rise of GenAI as a shift from crafting software, by hand, into creating an industrial process which advances software technology. This is a key way to understand the massive technological shift that we are seeing. More money, more power, more data, and more compute = technological progress. And through this lens it is easy to account for the “$600B AI Investment question”, not to mention the difference between AI hype and the actual utility of today’s solutions.

What do you think? Does AI have nat sec implications? And how should we be helping policy makers work through these questions? #genai #natsec #ai

Related Questions

What inspired Scott Persinger to start Supercog AI?
How did Scott Persinger's experience at Stripe influence his approach at Supercog AI?
What are the main challenges Scott Persinger faces as CEO of Supercog AI?
How does Supercog AI's technology differ from other AI integration solutions?
What is the future roadmap for Supercog AI according to Scott Persinger?
Scott Persinger
Scott Persinger, photo 1
Scott Persinger, photo 2
Add to my network

Location

San Francisco, California, United States