AI Agents for Negotiation Intelligence

In this week's podcast, I demonstrate how to build custom AI agents that analyze LinkedIn profiles for personality assessment and negotiation strategy development - potentially replacing paid tools with tailored solutions.

If you enjoyed this episode, please leave a review and check out our website: www.negoai.ai

I welcome any suggestions, questions, or comments at yrana@negoai.ai

--------------------------------------------------------------------------------------------------
Copyright Disclaimer Under Section 107 of the Copyright Act 1976, allowance is made for "fair use" for purposes such as criticism, comment, news reporting, teaching, scholarship, and research. Fair use is a use permitted by copyright statute that might otherwise be infringing.

Transcript

I've been using an AI app called Humantic for some time now. As you can see from my LinkedIn profile, it provides personality compatibility based on the DiSC profile of a person.

Humantic labels me and provides information about my educational level, experience, and descriptors like "ROI-driven," "precise," and "objective thinker." This is my DiSC profile, and the tool also suggests the best strategies for cold calling, email writing, qualifying or prospecting, and negotiating and closing.

It even has advanced insights like the OCEAN profile that you can see here. This is certainly a useful tool, but it's just a starting point when you're negotiating with somebody. So be flexible because, remember, it always gets information from the LinkedIn profile, and the LinkedIn profile may not match the real-life person.

Creating a Custom Agent

I wanted to create an agent that could replicate what Humantic does and potentially provide additional analysis, such as the Thomas-Kilmann conflict management style of a person.

I used Cassydi AI as my tool to work on agents and workflows. Here you can see I have four agents:

1. The trigger - a manual trigger that takes the LinkedIn profile of a person
2. The scraper - a built-in template in Cassydi that, when provided a LinkedIn profile, will scrape it and gather all the information
3. Behavioral agent number one - I've set the temperature to make it very deterministic. Remember, temperature ranges between zero and one: one is very creative, zero is very deterministic.
4. Agent number two - gets information from agent number one and is also very deterministic
5. Agent number three - I provided slightly more creativity to this agent because it's the one that will provide the final output

Agent Functions

Let me go through each of these agents to show you better how they work.

Agent number one gets the information from the LinkedIn scraper. It pre-scrapes the data from an individual's LinkedIn profile and identifies extra key information. It also cleans and pre-processes the text by applying natural language processing techniques. Essentially, it prepares the data for further analysis by agent number two.

Agent number two is the data science and psychometrical profiler specialist. It analyzes the pre-processed data provided by agent one and applies statistical and psychometrical models to generate detailed psychological profiles.

Agent number three is the strategist and report generation specialist. It uses the detailed psychological profiles provided by agent two to craft actionable sales and negotiation strategies. It also structures the content into a cohesive report.

Testing the Workflow

Let's go back to our workflow and insert my LinkedIn profile to see what it produces.

The workflow is now running. First, the built-in template in Cassydi will scrape my LinkedIn profile, providing all the information to agent number one. It will pre-process the information and apply natural language processing models before sending this information to agent number two. Agent number two will apply statistical and psychometrical models to the information, and that information will be fed into behavioral agent number three, which will provide us with the output.

Results and Analysis

The output includes my name, my role, and a profile confidence score.

Interestingly, the DiSC profile says I'm dominant and conscientious, which is quite in line with Humantic's assessment, which described me as more conscientious.

The report also includes my Big Five Personality Traits. But what's very interesting for me is the Thomas-Kilmann conflict management style, which identifies me as "collaborating with competitive tendencies."

The report further advises how to engage with me: leverage my strategic, innovative mindset by presenting complex negotiation challenges. It also details my communication style, negotiation strategy, and provides final insights that serve as a summary and implementation guidance.

I included a disclaimer because this is preliminary work, and the profile is generated based on LinkedIn data using AI analysis, so accuracy cannot be guaranteed.

Conclusion

This result is not very far from what Humantic provides us. Again, we are just in a preliminary phase, but as you can see, we can build agents that can provide us with very useful information about the other party before a negotiation.

Thanks for listening, and please subscribe to our newsletter at NegoAI.ai. Thank you.