This is a fascinating project! It's exciting to see the application of multi-agent simulations to a real-world problem like optimizing marketing messages. Your approach of grounding the AI personas in public data and focusing on social influence is very insightful.
I recently authored a paper on a related topic that might be of interest, particularly concerning your challenges with accuracy and long-term simulation stability. Our work, "SALM: A Multi-Agent Framework for Language Model-Driven Social Network Simulation," introduces a novel framework for integrating language models into social network simulations.
I specifically focused on achieving long-term temporal stability in multi-agent scenarios.[2] A couple of our key contributions could be relevant to your work:
I developed a hierarchical prompting architecture that enables stable simulations beyond 4,000 timesteps while significantly reducing token usage.[3]
To address memory growth and maintain agent consistency, I implemented an attention-based memory system that achieves high cache hit rates with sub-linear memory growth.
I also established formal bounds on personality stability, which could be a useful concept for ensuring the accuracy and predictability of your AI personas over time.
Our validation against SNAP ego networks demonstrated the capability of our framework to model long-term social phenomena with empirically validated behavioral fidelity.[3]
It seems like there's a lot of synergy between our research and your product. I'd be very interested to hear your thoughts on our paper and explore if any of our findings could be beneficial for the continued development of Artificial Societies.
Working on a resume editing tool for rapidly applying to a large number of jobs. The basic idea being that resumes should be tailored to the job being applied to. In this tool, you put in your resume, add in other details regarding your background, projects and other things that you might not put in the resume. You do this once.
Now you paste in the job description, we use an llm to generate a new resume grounded in the details you added earlier. You get a pdf that you can submit to the job application.
I have worked on social dynamics on social media particularly how why certain kinds of content/people become viral. To that end, I have been working on a large-scale simulation that simulates twitter. It pulls data from a random distribution, generates users with specific interest vectors. its modelled as a Agent based modelling with the users as agents, posting every timestep, reading content posted by others, recommended to them by a recommendation engine. all of this this then generates data for research.
Love the work you're doing. I'm looking for a part-time role. Based out of NJ, USA. I haven't worked with Crystal but have extensive Ruby experience. Sent you an email!
Location: New Jersey, US
Remote: Yes
Willing to relocate: No
Technologies: Python, JS, Typescript, Node, SQL (Postgres, MSSQL), Azure Functions, Docker, R, Apache JMeter, Ruby on Rails,
LinkedIn: https://www.linkedin.com/in/arkokoley/
Email: gaurav at koley.in
Website: gaurav.koley.in
Hi all, I'm an ex-Microsoft SWE currently working through my PhD in Data Science at Boston University. In the past I have worked on numerous projects with Large Enterprises and small startups. I am a one man machine, capable of working on Front-end, backend and any devops aspects of any business. I'm also a Data Scientist in training so interested in data-sciency roles too!
I used to do that with QuickRide in Bangalore. Very popular and fairly simple to use. As a rider, you can put up your origin and destination points. You are shown people with cars who have a high overlap with your origin-destination route and you can send them requests to join.
Usually 1/4th the price of a cab and the app has information about where each rider/car owner works (mostly MNCs), that works as a trust factor.
I used to work for Microsoft and our team's morning rant (on Whatsapp) would start with MS Teams not working. Often meetings would be stopped because Teams hung up for the person who was presenting. And the Teams/Skype Infra team were the slowest in the world to respond to any requests!
I had more or less the same experience, but TBH MS uses internally the latest unreleased version of applications so those problems are partially to be expected and act as dog fooding/user testing
haha, this explains. somehow the Teams team does not care, or the product/ops team is organized in a rigid way, hard to make some changes even the dev feels the inconvenience.
The article mistakenly calls the Haken Kreuz as Swastika, confusing the Hindu symbol for prosperity with the Nazi Haken Kreuz. This should be corrected.
I recently authored a paper on a related topic that might be of interest, particularly concerning your challenges with accuracy and long-term simulation stability. Our work, "SALM: A Multi-Agent Framework for Language Model-Driven Social Network Simulation," introduces a novel framework for integrating language models into social network simulations.
I specifically focused on achieving long-term temporal stability in multi-agent scenarios.[2] A couple of our key contributions could be relevant to your work:
I developed a hierarchical prompting architecture that enables stable simulations beyond 4,000 timesteps while significantly reducing token usage.[3]
To address memory growth and maintain agent consistency, I implemented an attention-based memory system that achieves high cache hit rates with sub-linear memory growth.
I also established formal bounds on personality stability, which could be a useful concept for ensuring the accuracy and predictability of your AI personas over time.
Our validation against SNAP ego networks demonstrated the capability of our framework to model long-term social phenomena with empirically validated behavioral fidelity.[3]
It seems like there's a lot of synergy between our research and your product. I'd be very interested to hear your thoughts on our paper and explore if any of our findings could be beneficial for the continued development of Artificial Societies.
Here's the link to my paper: https://arxiv.org/abs/2505.09081