Tyrin-Ian Todd
Founder of Modaic, Verification & alignment infra for AI decisions. MIT / BU team, design partnerships with Accenture, Dropbox, and Vercel.
Company
Tyrin-Ian Todd is listed as a founder of Modaic. Verification & alignment infra for AI decisions. MIT / BU team, design partnerships with Accenture, Dropbox, and Vercel. Speedrun 006 profile: 2 employees; located in San Francisco, California; tags: Infra, AI.
Founder traction evidence
Public posts and activity attributed directly to this founder.
- LinkedIn
Farouk Adeleke and I are proud to announce that Modaic has joined a16z speedrun.
Farouk Adeleke and I are proud to announce that Modaic has joined a16z speedrun.
- LinkedIn
Almost a year ago today I graduated from Massachusetts Institute of Technology with a degree in AI and Decision Making (course 6-4) and now I'm back on campus for Boston TECH WEEK by a16z!
Modaic (SR 006) is backed by Andreessen Horowitz a16z speedrun building exactly what my major was named after — AI for decision making.
- X
Today we’re launching @modaicdev @a16z @speedrun 006, the fastest way to render your judgement into reliable decision automation.
Today we’re launching @modaicdev @a16z @speedrun 006, the fastest way to render your judgement into reliable decision automation.
- X
Getting LLMs to Quantify Their Unknowns LLM judges are increasingly common among AI teams due their ability to automate decisions that require complex reasoning and analysis.
Getting LLMs to Quantify Their Unknowns LLM judges are increasingly common among AI teams due their ability to automate decisions that require complex reasoning and analysis. Pairing their reasoning ability with calibrated confidence scores
- X
How you you get LLMs to quantify their uncertainty? I wrote a blog comparing the top blackbox with logit probes on various LLM judgement benchmarks Super interesting findings on how LLMs express uncertainty under the surface.
How you you get LLMs to quantify their uncertainty? I wrote a blog comparing the top blackbox with logit probes on various LLM judgement benchmarks Super interesting findings on how LLMs express uncertainty under the surface. https://t.co/Yf2xTCkOEx
- X
git mogged
git mogged
- X
@modaicdev uses SOTA techniques from mechanistic interpretability to look inside the LLM and extract calibrated confidence scores for ev...
@modaicdev uses SOTA techniques from mechanistic interpretability to look inside the LLM and extract calibrated confidence scores for every decision. Only low confidence decisions are routed to humans for review.
- X
What ever happened to A2A?
What ever happened to A2A?