🌴⚡️Have your agent call my agent

The Nomad Technologist
Get smarter about AI, tech and business in 5 minutes or less.
The technical enablement newsletter for globally distributed teams.

🌴 The Nomad Technologist ⚡️ 

While everyone's debating whether AI will replace us, Kobe and I have been experimenting with these tools and researching how they're transforming the systems we architect. Today we're diving deep into how agentic AI, LLMs, and advanced analytics are reshaping customer experience. Think Agent to Agent and Model Context Protocol.

🤖 Agentic AI: Beyond the Chatbot Hype

Forget everything you think you know about customer service bots. Agentic AI isn't just responding to queries. It's making autonomous decisions and pursuing goals with minimal human oversight.

(think: AI that can actually solve problems end to end, not just answer FAQs).

Here's the shift that caught my attention: These systems don't just process data; they take action. They can handle entire customer journeys from inquiry to resolution, make complex decisions, and adapt in real time. Over 60% of executives surveyed believe this will fundamentally improve customer satisfaction.

What this means for your stack (and your business):

• Your APIs need to handle more complex, multi-step AI interactions (systems that can remember context across multiple requests)

• Consider building systems that can gracefully hand off between AI and human agents (seamless escalation paths)

• Think about data architecture that supports real time decision-making at scale (instant access to customer history, preferences, and context)

Learning note: As I explore this space myself, I'm seeing that starting small with clear goals and guardrails makes sense. Thus far it seems agents are best implemented when broken up into individual task specialists and chaining multiple “specialist” agents together to make a system.

🌴 The Nomad Technologist ⚡️ 

📊 Hyper Personalization That Actually Works

Here's a stat that should make every product engineer spit coffee on their keyboard: hyper-personalized experiences generate up to 40% more revenue than generic ones. 

Now we’re not talking about basic "Hi [FIRST_NAME]" email templates.

Modern AI systems are analyzing:

• Real time behavioral data (what users are doing right now on your site/app)

• Predictive patterns from historical interactions (spotting trends before they happen)

• Cross channel activity synthesis (connecting mobile, web, email, and in store behavior)

• Intent detection and sentiment analysis (understanding not just what users do, but why)

The insight that's reshaping how I think about product architecture: If you're building customer facing products, your data pipeline architecture is now your competitive moat. The companies crushing it are those who can integrate omnichannel data in real time and serve personalized experiences at the API level (basically: your ability to know your customer across every touchpoint instantly).

Industry applications getting results:

• Retail: Dynamic product recommendations based on real time browsing (Amazon showing you exactly what you want, when you want it)

• FinTech: Personalized financial products based on spending patterns (credit cards that adapt to your lifestyle)

• Healthcare: Treatment plans tailored to genetic and lifestyle data (medicine that's actually designed for you)

🌴 The Nomad Technologist ⚡️ 

🗣️ LLMs in Production: Beyond the Demo

The gap between “cool ChatGPT demo bro” and a production ready customer experience is massive. Here's what I’ve observed is actually working in the real world:

Natural Language Understanding + Generation is creating conversations that feel genuinely helpful. 51% of consumers now prefer interacting with well designed AI when they want immediate service (faster than waiting on hold).

The engineering reality (and business opportunity):

• Your LLM needs access to real time, company specific data (RAG architecture)

• Multilingual support is table stakes (50+ languages expected, especially for global teams)

• Context management across multi-turn conversations is non trivial (remembering what was said 5 messages ago)

  
  Nomad Technologist Tip: A proven way to boost adoption of AI agents in customer service is to delay revealing it's an AI. Let the agent greet the user, listen fully, and respond thoughtfully. This will allow the customer to be impressed by the thoughtful response of the customer service AI Agent. At that point it can mention that is is an AI agent briefly while focusing on the task at hand (having earned trust through its response). Ensure the human still has the ability to be transferred or escalated to a human if they choose. For my fellow rabbit-hole divers you can find the episode here.
  The Polyglot Software Development Podcast
  
  Only after delivering value should it disclose it's AI. This builds trust through performance, not labels.

🌴 The Nomad Technologist ⚡️ 

🔗 Omnichannel Integration: The Data Engineering Challenge

Customers expect seamless transitions between web, mobile, phone, and in store interactions (starting a question on chat, continuing over email, finishing on a phone call). This isn't a UX problem, it's a data architecture problem.

The technical challenge (and why it matters for your business):

• real time data synchronization across channels (every system talking to each other instantly)

• Context preservation during channel switches (no sorry can you repeat that? moments)

• Customer preference learning and optimization (AI figuring out how each customer likes to communicate)

• Emotional state tracking across touchpoints (detecting frustration before it escalates)

Something Kobe and I are seeing consistently: Companies are building unified customer data platforms that can process and interpret behavior patterns across channels in real time, then serve that context to any touchpoint instantly (think one source of truth for every customer interaction).

🌴 The Nomad Technologist ⚡️ 

🔗 Omnichannel Integration: The Data Engineering Challenge

Customers expect seamless transitions between web, mobile, phone, and in store interactions (starting a question on chat, continuing over email, finishing on a phone call). This isn't a UX problem, it's a data architecture problem.

The technical challenge (and why it matters for your business):

• real time data synchronization across channels (every system talking to each other instantly)

• Context preservation during channel switches (no sorry can you repeat that? moments)

• Customer preference learning and optimization (AI figuring out how each customer likes to communicate)

• Emotional state tracking across touchpoints (detecting frustration before it escalates)

Something Kobe and I are seeing consistently: Companies are building unified customer data platforms that can process and interpret behavior patterns across channels in real time, then serve that context to any touchpoint instantly (think one source of truth for every customer interaction).

🌴 The Nomad Technologist ⚡️ 

🧠 Emotion AI: The Sentiment Analysis Evolution

Emotion AI is analyzing voice, text, and facial expressions in real time to detect frustration, confusion, satisfaction, or urgency. This isn't just "positive vs negative” it's emotional intelligence at scale (AI that can read the room).

The engineering opportunity (and customer experience win):

• real time emotional analysis APIs (detecting mood in milliseconds)

• Automated escalation based on emotional indicators (routing angry customers to your best agents)

• Predictive emotional analytics for proactive support (stopping problems before they start)

• Integration with existing customer journey orchestration (making every touchpoint emotionally aware)

Companies implementing this report improved retention rates and stronger customer loyalty. The key is building systems that can detect emotional trajectories and intervene proactively (like a good bartender who knows when someone's having a bad day).

🌴 The Nomad Technologist ⚡️ 

🔍 RAG: The Knowledge Architecture That Actually Works

Retrieval-Augmented Generation is solving the hallucination problem by grounding AI responses in your actual knowledge base. Instead of AI making up plausible sounding answers, RAG systems retrieve real information first, then generate contextually relevant responses (AI that actually knows your business, not just generic internet knowledge).

Why I'm getting excited about this for solo founders: This requires sophisticated information retrieval systems, semantic search capabilities, and careful integration with your existing knowledge management infrastructure, but the payoff in customer experience could be massive.

What this means practically:

• Your knowledge bases become dynamic, queryable resources (documentation that talks back)

• real time information retrieval combined with natural language generation (instant, accurate answers)

• Hyper personalized responses based on customer specific context (answers tailored to each situation)

• Empowers support agents with instant access to comprehensive information (everyone becomes an expert)

🌴 The Nomad Technologist ⚡️ 

⚖️ Ethical AI: The Trust Infrastructure

As AI becomes more prevalent in customer interactions, ethical implementation isn't just nice to have. It's business critical infrastructure (customers need to trust your AI to trust your business).

Key technical considerations:
(and why they matter for your brand, customers or company)

• Algorithmic fairness auditing and bias detection (making sure AI treats everyone fairly)

• Explainable AI for significant customer decisions (customers deserve to know why AI made certain choices)

• Privacy by design data governance frameworks (protecting customer data from day one)

• Customer control and opt out mechanisms (always giving customers choice)

What I'm learning from studying companies doing this right: They're prioritizing ethical AI not just for compliance, but because it builds stronger customer trust and more sustainable competitive advantages. This means building transparency and fairness into your systems from day one, not as an afterthought.

🌴 The Nomad Technologist ⚡️ 

🚀 What This Means for Your Next Sprint

If you're building customer facing products:

1. Audit your data architecture: Can you serve real time, personalized experiences? (Do your systems talk to each other?)

2. Evaluate RAG implementation: How could your knowledge base become a dynamic AI resource? (Can your documentation answer questions automatically?)

3. Plan for emotional intelligence: Where could sentiment analysis improve your user experience? (Where do customers get frustrated?)

4. Design for transparency: How will you explain AI decisions to users? (Can you show your work?)

If you're on distributed teams (or managing them):

• These AI systems require sophisticated orchestration and real time data processing (lots of moving parts working together)

• Consider how your microservices architecture can support AI driven personalization (each service needs to be AI aware)

• Think about API design for seamless human to AI handoffs (smooth transitions between automated and human help)

🌴 The Nomad Technologist ⚡️ 

🥇 The Gold Medal

The companies winning with AI in customer experience aren't just implementing tools. They're rethinking their entire data and system architecture to support intelligent, autonomous decision making at scale.

Two things to remember from todays newsletter: 
Agent to Agent
Model Context Protocol

The opportunity is massive. Improved. customer satisfaction, reduced operational costs, and increased revenue through personalization. But the technical execution complexity is real (this isn't a weekend hackathon project).

The Nomad Technologist Challenge: Pick one area where your current customer experience creates friction, and prototype how AI could make it seamless. Start small, measure everything, and scale what works.

Think about some triggers that would result in an the agent escalating the call to a human.

For example the customer is upset, or the AI is hallucinating and unable to properly service the customers inquiries.

The future isn't human vs. AI

It's building systems where both work together to create experiences that neither could deliver alone.

Unless they kill us. There’s that too.
 
Cheers to your success.
Chase & Kobe 👨‍💻🐶 

🌴 The Nomad Technologist ⚡️ 

🌴 The Nomad Technologist ⚡️