Vellum is coming to the AI Engineering World's Fair in SF. Come visit our booth and get a live demo!

MCP: The Hype vs. Reality

LLMs are stepping outside the sandbox. Should you let them?

5 min
Written by
Reviewed by
No items found.

We’ve been slowly handing more control to LLMs.

First, we let them choose what to say (prompting). Then, we let them choose how to act (function calling). Next came more advanced orchestration (memory, RAG, multi-step reasoning), so they could plan and refine before they generate an output.

But all of that has stayed inside the workflow.

What if we let them go further? What if an LLM could dynamically discover and use external tools and resources through one single interface? Your CRM, your API stack, your docs in Notion, your messages in Slack?

Model Context Protocol (MCP) is a first step into that world.

An open standard designed to fundamentally change how LLMs connect to the outside world.

It’s a bold vision, but is it ready for real-world use?

What is the Model Context Protocol (MCP)?

The Model Context Protocol (MCP) is an open protocol specification, introduced and open-sourced by Anthropic in November 2024, that standardizes the communication between LLM-powered applications and external data sources or tools.

Think of MCP as a universal adapter or a "USB-C port" for AI applications.

Just as USB-C provides a standard way to connect diverse peripherals to a computer, with power and data flowing in both directions, MCP offers a unified way for any compliant AI application (an MCP Host or Client) to interact with any compliant external resource (an MCP Server).

MCP’s main goal is to replace messy, fragile integrations with one solid, secure standard, so developers can focus on building their app, not wiring it up to every tool one by one.

It borrows the concept of standardized interactions from APIs and LSP. Like APIs standardize communication between apps, and LSP standardizes communication between IDEs and languages, MCP standardizes how AI models (like chatbots or AI agents) interact with external tools, data, and resources.

Similarities between APIs, LSP and MCP

Why is MCP winning right now?

Last month, Swyx, an independent AI engineer said: “It’s fair to say that MCP has captured enough critical mass and momentum right now that it is already the presumptive winner of the 2023-2025 “agent open standard” wars.”

This is generally seeming like the situation we’re in, as we saw MCP trending faster than any other framework/tool on the market.

Is this just another hype cycle?

What signs can we look at to determine if it’s more than that?

Well, the biggest benefit is that you only need one MCP Client to work with any MCP server. No custom integrations required.

This means AI applications can easily connect to new tools without needing new custom connection each time. By using MCP, AI systems become more efficient and easier to scale.

Plus, it’s backed by a growing ecosystem of open-source servers and SDKs across major languages.

Arguments against MCP

MCP is a promising but very new (just three months old). We usually recommend waiting how new tech evolves before investing too much time into it.

Although developers love the new shiny thing, there have been a few contra arguments of why this protocol might not be the best solution.

Here's where hype crashes into real-world limits:

1/ MCP is currently a stateful protocol

Integrating MCP with serverless AI systems is challenging because MCP relies on long-lived, stateful connections. And developers like serverless architecture for AI systems because it’s cheaper, and easier to scale.

Even the creator, Justin Spahr-Summers is aware of this limit. In their forum he explicitly says: “Deploying to a Platform-as-a-Service is really nice and convenient as a developer, so not being very compatible with this model creates an impediment to broader MCP adoption.”

2/ Reinventing the wheel

MCP aims to simplify how AI agents connect to tools, but most companies already use REST APIs for this. Asking server devs to build new endpoints and run long-lived servers just for MCP feels unrealistic.

Some developers are leaning towards using more lean implementations like agents-json, that formally describes contracts for API and agent interactions, built on top of the OpenAPI standard.

3/ Controlling AI systems in production is already hard

Managing AI in production is already very unreliable and hard. LLMs don't "understand" context in the human sense, and their interpretation of context is highly sensitive to prompt phrasing and bias. Integrating a protocol that can decide on its own on what tools/resources it can use to achieve a task is going to be very hard to evaluate and manage. So, why do we want to add more uncertainty with something that is already a high-risk?

I get the need to make space for new tech, but if you're building for production right now, it might be smarter to wait and see how this cycle plays out.

Advice for companies serious about AI

At Vellum, we’re focused on making it easier for teams to build AI into production systems. That means helping them manage complexity, not add to it.

MCP is exciting because it promises a cleaner, standardized way for LLMs to work across tools, APIs, and workflows. It aligns with our goal of giving teams building with AI a more modular, reusable stack.

But while the vision is strong, MCP is still early.

Most enterprises don’t need full agentic autonomy across systems just yet. What they do need is control, observability, and a clear path to production. That’s why our advice today is to keep an eye on MCP, experiment if you’re curious, but don’t rebuild your stack for it just yet. The better move right now is to use battle-tested methods (like REST and webhooks) while the ecosystem matures.

If MCP lives up to its promise, we’ll be ready to support it when the time is right. And if you’re building toward that future, we’re here to help -- book a call with our experts here.

ABOUT THE AUTHOR
Anita Kirkovska
Founding Growth Lead

An AI expert with a strong ML background, specializing in GenAI and LLM education. A former Fulbright scholar, she leads Growth and Education at Vellum, helping companies build and scale AI products. She conducts LLM evaluations and writes extensively on AI best practices, empowering business leaders to drive effective AI adoption.

ABOUT THE reviewer

No items found.
lAST UPDATED
Apr 9, 2025
share post
Expert verified
Related Posts
LLM basics
October 10, 2025
7 min
The Best AI Workflow Builders for Automating Business Processes
LLM basics
October 7, 2025
8 min
The Complete Guide to No‑Code AI Workflow Automation Tools
All
October 6, 2025
6 min
OpenAI's Agent Builder Explained
Product Updates
October 1, 2025
7
Vellum Product Update | September
Guides
October 6, 2025
15
A practical guide to AI automation
LLM basics
September 25, 2025
8 min
Top Low-code AI Agent Platforms for Product Managers
The Best AI Tips — Direct To Your Inbox

Latest AI news, tips, and techniques

Specific tips for Your AI use cases

No spam

Oops! Something went wrong while submitting the form.

Each issue is packed with valuable resources, tools, and insights that help us stay ahead in AI development. We've discovered strategies and frameworks that boosted our efficiency by 30%, making it a must-read for anyone in the field.

Marina Trajkovska
Head of Engineering

This is just a great newsletter. The content is so helpful, even when I’m busy I read them.

Jeremy Hicks
Solutions Architect

Experiment, Evaluate, Deploy, Repeat.

AI development doesn’t end once you've defined your system. Learn how Vellum helps you manage the entire AI development lifecycle.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Build AI agents in minutes with Vellum
Build agents that take on the busywork and free up hundreds of hours. No coding needed, just start creating.

General CTA component, Use {{general-cta}}

Build AI agents in minutes with Vellum
Build agents that take on the busywork and free up hundreds of hours. No coding needed, just start creating.

General CTA component  [For enterprise], Use {{general-cta-enterprise}}

The best AI agent platform for enterprises
Production-grade rigor in one platform: prompt builder, agent sandbox, and built-in evals and monitoring so your whole org can go AI native.

[Dynamic] Ebook CTA component using the Ebook CMS filtered by name of ebook.
Use {{ebook-cta}} and add a Ebook reference in the article

Thank you!
Your submission has been received!
Oops! Something went wrong while submitting the form.
Button Text

LLM leaderboard CTA component. Use {{llm-cta}}

Check our LLM leaderboard
Compare all open-source and proprietary model across different tasks like coding, math, reasoning and others.

Case study CTA component (ROI)

40% cost reduction on AI investment
Learn how Drata’s team uses Vellum and moves fast with AI initiatives, without sacrificing accuracy and security.

Case study CTA component (cutting eng overhead) = {{coursemojo-cta}}

6+ months on engineering time saved
Learn how CourseMojo uses Vellum to enable their domain experts to collaborate on AI initiatives, reaching 10x of business growth without expanding the engineering team.

Case study CTA component (Time to value) = {{time-cta}}

100x faster time to deployment for AI agents
See how RelyHealth uses Vellum to deliver hundreds of custom healthcare agents with the speed customers expect and the reliability healthcare demands.

[Dynamic] Guide CTA component using Blog Post CMS, filtering on Guides’ names

100x faster time to deployment for AI agents
See how RelyHealth uses Vellum to deliver hundreds of custom healthcare agents with the speed customers expect and the reliability healthcare demands.
New CTA
Sorts the trigger and email categories

Dynamic template box for healthcare, Use {{healthcare}}

Start with some of these healthcare examples

SOAP Note Generation Agent
Extract subjective and objective info, assess and output a treatment plan.
Population health insights reporter
Combine healthcare sources and structure data for population health management.

Dynamic template box for insurance, Use {{insurance}}

Start with some of these insurance examples

Agent that summarizes lengthy reports (PDF -> Summary)
Summarize all kinds of PDFs into easily digestible summaries.
Insurance claims automation agent
Collect and analyze claim information, assess risk and verify policy details.
AI agent for claims review
Review healthcare claims, detect anomalies and benchmark pricing.

Dynamic template box for eCommerce, Use {{ecommerce}}

Start with some of these eCommerce examples

E-commerce shopping agent
Check order status, manage shopping carts and process returns.

Dynamic template box for Marketing, Use {{marketing}}

Start with some of these marketing examples

LinkedIn Content Planning Agent
Create a 30-day Linkedin content plan based on your goals and target audience.
ReAct agent for web search and page scraping
Gather information from the internet and provide responses with embedded citations.

Dynamic template box for Legal, Use {{legal}}

Start with some of these legal examples

PDF Data Extraction to CSV
Extract unstructured data (PDF) into a structured format (CSV).
Legal document processing agent
Process long and complex legal documents and generate legal research memorandum.

Dynamic template box for Supply Chain/Logistics, Use {{supply}}

Start with some of these supply chain examples

Risk assessment agent for supply chain operations
Comprehensive risk assessment for suppliers based on various data inputs.

Dynamic template box for Edtech, Use {{edtech}}

Start with some of these edtech examples

Turn LinkedIn Posts into Articles and Push to Notion
Convert your best Linkedin posts into long form content.

Dynamic template box for Compliance, Use {{compliance}}

Start with some of these compliance examples

No items found.

Dynamic template box for Customer Support, Use {{customer}}

Start with some of these customer support examples

Q&A RAG Chatbot with Cohere reranking
Trust Center RAG Chatbot
Read from a vector database, and instantly answer questions about your security policies.

Template box, 2 random templates, Use {{templates}}

Start with some of these agents

Research agent for sales demos
Company research based on Linkedin and public data as a prep for sales demo.
Agent that summarizes lengthy reports (PDF -> Summary)
Summarize all kinds of PDFs into easily digestible summaries.

Template box, 6 random templates, Use {{templates-plus}}

Build AI agents in minutes

Legal RAG chatbot
Chatbot that provides answers based on user queries and legal documents.
Clinical trial matchmaker
Match patients to relevant clinical trials based on EHR.
E-commerce shopping agent
Check order status, manage shopping carts and process returns.
Legal contract review AI agent
Asses legal contracts and check for required classes, asses risk and generate report.
Financial Statement Review Workflow
Extract and review financial statements and their corresponding footnotes from SEC 10-K filings.
AI legal research agent
Comprehensive legal research memo based on research question, jurisdiction and date range.

Build AI agents in minutes for

{{industry_name}}

Clinical trial matchmaker
Match patients to relevant clinical trials based on EHR.
Prior authorization navigator
Automate the prior authorization process for medical claims.
Population health insights reporter
Combine healthcare sources and structure data for population health management.
Legal document processing agent
Process long and complex legal documents and generate legal research memorandum.
Legal contract review AI agent
Asses legal contracts and check for required classes, asses risk and generate report.
Legal RAG chatbot
Chatbot that provides answers based on user queries and legal documents.

Case study results overview (usually added at top of case study)

What we did:

1-click

This is some text inside of a div block.

28,000+

Separate vector databases managed per tenant.

100+

Real-world eval tests run before every release.