There’s a lot of talk right now about “AI” in real estate. But too often, that talk gets wrapped in jargon and hype. So let’s cut through the noise. As we build out custom AI solutions for clients, we are gaining deeper understanding of the importance of AI model selection. Here are some recent learnings from SLM usage.

If you’re a real estate exec, running a brokerage, team, MLS, or proptech company, and someone mentions “SLMs” and “LLMs,” here’s what they’re really talking about:

SLMs are smaller, faster, cheaper AI models that can be highly customizable.

Think nimble.

LLMs are bigger, broader in knowledge, and more expensive AI models, and are only configurable through prompts and the information context you feed them.

Think powerful.

It is not necessary to understand how the engine works. However, it is important to know when to choose a hybrid versus a truck. In this analogy, a hybrid represents an SLM, which is efficient and suitable for specific, streamlined tasks. A truck represents an LLM, robust and capable of handling more complex, broader challenges with more power.

So, what’s the real difference?

Let’s start simple.

If the goal is to do one specific thing well, such as automating listing input or generating lead responses, an SLM might be required to do the job.

But if you’re building a more complex tool, like a smart assistant that understands contracts, listing history, client tone, and market shifts, an LLM might serve you better.

Models are classified as either open-source or paid/proprietary models. The following are key points about the differences here.

Open-Source Models (LLMs & SLMs)

• Examples of these models are Meta LLaMA, Mistral, Falcon, Gemma, DeepSeek, and T5.
• Platforms and libraries that support open-source models are Hugging Face, PyTorch, and TensorFlow.
• The benefits of open-source models include full transparency, extensive customization/fine-tuning, no vendor lock-in, community-driven development, and lower costs (paying for infrastructure).
• Downsides include the need for technical skills to deploy, potential lack of official support, and the resources to manage the infrastructure on which they are hosted.

Paid/Proprietary Models (LLMs)

• Examples of these models are OpenAI’s GPT-4o, Anthropic’s Claude, and Google’s Gemini (API versions).
• Platforms supporting them include the OpenAI API, the Anthropic API, and Google AI Platform.
• The benefits are state-of-the-art performance, user-friendly interfaces, dedicated support, and managed infrastructure (pay-per-token).
• Downsides are higher recurring costs, data privacy concerns (the possibility of sending data to the vendor), and limited control over model architecture.

The strategic decision tree

Before you spend a dime on AI, slow down and ask one hard question:

What are we actually trying to solve?

Not every problem needs a giant model. Some problems are better tackled with a fast, lightweight tool that does one thing well. Others require a more powerful system that can juggle nuance, compliance, and volume.

Consider cost, speed, privacy, and complexity as the main factors when deciding on a model. Here’s a simple decision tree to help an organization decide when to use an SLM, when to use an LLM, and when not to bother with either.

• Is the data you’re working with sensitive or regulated?
  • If yes, and you need to keep it local (e.g., on-prem or on device), use an SLM.
• Is cost or speed a major constraint?
  • Again, that’s a point for SLMs.
• Do you want a fast launch and don’t mind using a cloud API?
  • That’s a green light for LLMs, services like OpenAI’s GPT or Google’s Gemini.
• Is your problem about keeping up with knowledge (market stats, trends, legal docs)?
  • Don’t fine-tune anything. Use RAG (retrieval-augmented generation), a process that improves model responses by retrieving relevant information from a database. It’s cheaper and better for updating facts.
• Is your problem about control (tone, format, behavior)?
  • That’s where fine-tuning (especially on SLMs) can shine.
• Are you trying to add a new capability, like interpreting local MLS policy or translating listing slang?
  • SLMs with domain-specific fine-tuning may be your best bet.

Real estate examples

The following examples can provide insight into where and when to use either model.

When SLMs are enough

A tool designed to analyze agent performance against proprietary business and agent data. Each real estate brokerage uses unique terminology and performance metrics. An SLM offers the nimbleness and flexibility required to specialize in data collection and analysis, supporting managerial decision-making.

An SLM can be configured as a listing input assistant that specializes in managing data entry across multiple MLSs. Once set up for the specific requirements of each system, the SLM can efficiently handle and automate the process of entering listing data into various MLS platforms based on the brokerage’s and agents’ participation. This approach allows brokerages and agents to streamline operations, reduce repetitive manual work, and ensure consistency and accuracy of listing information across all relevant databases.

Localized lead response bots can be fine-tuned not only for a single market, but also for the specific agent assigned to a lead. By customizing the bot to reflect the agent’s style, preferences, and communication habits, these systems can help maintain consistent, personalized engagement with potential clients. As a result, the response bot acts as an extension of the agent, assisting in keeping the agent in touch with customers and ensuring timely, relevant follow-ups throughout the client journey.

When you need LLMs

A cross-market consumer chatbot can be designed to communicate fluently in multiple languages and possess in-depth knowledge of various loan programs. Beyond its foundational multilingual capabilities, this type of chatbot can be further customized to address the unique requirements and preferences of different markets and the specific needs of individual users. This level of adaptability makes the chatbot a valuable resource for diverse client bases seeking assistance across regions and languages.

A writing tool that drafts listing descriptions with style matching and compliance baked in. Modern AI platforms like ChatGPT’s GPTs, Claude Skills, and Google’s Gemini can take this even further. These advanced models not only generate content but can also be customized to reflect the unique voice of a brokerage or individual agent.

Anything that connects deeply with dozens of tools via API (CRM, TMS, MLS, CMA tools). Increasingly, advanced AI models are leveraging not just traditional APIs but also Model Context Protocol (MCP) Servers to access and incorporate additional data sources into their workflows. By utilizing MCP Servers, these systems can dynamically pull in relevant information from a wide variety of structured and unstructured data repositories, further enriching their responses.

One last note on cost

Fine-tuning a big model (LLM) isn’t just expensive once, it becomes a recurring investment. You retrain it every time the market shifts, laws change, or your tone needs an update.

For example, implementing LLMs like GPT-4 can range from thousands to millions of dollars annually, depending on scale and usage.

SLMs, on the other hand, are cheap enough to experiment with. You can tune them fast and often, or run multiple versions for different teams or create A/B testing scenarios. Costs for SLMs are significantly lower, often in the range of hundreds to a few thousand dollars per year.

Final thoughts

You don’t need to bet the farm on the biggest model.

If you’re clear about your problem, cost, speed, privacy, or control, the choice between SLMs and LLMs becomes obvious.

Start small. Pilot something. Let the model earn its keep.

Because in this market, even a well-placed tool that saves 10 minutes per agent per day can move the needle.

And that’s worth paying attention to.

There is a new class of models in town, Multimodal Language Models (MLMsj). These models have emerged to address the growing demand for handling more than just text. They can also process audio and video inputs.

This increased importance reflects the need for AI systems to interpret and generate responses across diverse media types, making them especially valuable for applications that require understanding and synthesizing information from multiple sources.

Stay tuned as we will explore these models in depth. If you need a consultant to help you with your AI strategy or AI development in real estate, we would love to talk to you.

The post SLMs vs LLMs in Real Estate AI Tools and Products appeared first on WAV Group Consulting.

The conversation about AI in real estate has moved past demos and experiments. We’re now entering a phase where the infrastructure underneath these systems matters as much as the tools themselves.

This month, the Linux Foundation announced the formation of the Agentic AI Foundation (AAIF), a new standards body designed to make AI agents interoperable, governable, and safe.

It brings together heavyweight members like AWS, Google, Microsoft, OpenAI, and Block, along with others such as IBM, Cisco, and Salesforce. These companies compete fiercely in the AI market, but under the AAIF banner, they’ve agreed to collaborate on common ground, open standards that everyone can build upon.

That kind of neutrality is exactly what the real estate industry needs.

Why Real Estate Should Pay Attention

For years, brokerages, MLS organizations, and proptech firms have built integrations in piecemeal ways, one vendor or API at a time. Each connection required custom engineering, and every update risked breaking the system. As a result, continuous management of high costs, fragile workflows, and an environment where innovation often depended on a single provider’s roadmap.

AAIF changes that dynamic by introducing shared, open protocols for AI agent development and orchestration. Its first three projects are industry-leading tools: Anthropic’s Model Context Protocol (MCP), Block’s goose framework, and OpenAI’s AGENTS.md.

A combination of products that define how AI agents connect to tools, data, and workflows, and how they should behave once they do. Together, they shift AI integration away from vendor-specific APIs toward a consistent, auditable standard.

For real estate technology leaders, this is a governance pivot. Similar to the impact of RESO’s data standards two decades ago.

Just as RESO standardized listing data across MLSs, MCP and its companion frameworks are standardizing how AI interacts with real estate data, CRMs, marketing systems, and transaction platforms. It’s an interoperability layer for the AI era.

The Integration Layer for AI

From a practical perspective, MCP allows a brokerage, MLS, or vendor platform to expose its capabilities in a structured, discoverable way. Instead of custom API endpoints or private integrations, an AI agent can query an MCP server to understand what operations are possible.

Think of MCP as a USB hub that connects your devices. Creating the ability to search listings, create CMAs, update transaction milestones, lead response generation, agent productivity coaching, or generate marketing assets.

This kind of design also introduces governable access. Brokerages and MLSs can define permissions, control context, and audit AI-driven actions (an overlooked requirement, but it is absolutely necessary). It aligns with data privacy and compliance requirements while still enabling automation and innovation.

goose and AGENTS.md Enabling Governance for the AI Era

The other two AAIF projects extend that governance idea into operations. “goose” provides a local-first framework for building structured and auditable AI workflows. A must for brokerages that want to automate tasks like lead routing, marketing setup, or compliance reviews without exposing sensitive data.

AGENTS.md plays a simpler but equally important role. This little file provides developers and organizations with a standard place to define rules and expectations for AI agents within a project.

In a real estate context, that could include brand guidelines, jurisdictional constraints, and data-handling policies, such as the “how we do things here” file for digital staff.

The broader impact is that AI can now move from being a set of disconnected pilot projects into a core part of brokerage operations. When standards exist, investment risk goes down. When governance is shared, trust goes up.

Building Confidence Through Open Governance

Real estate organizations can build with confidence that their AI integrations will last longer than a product cycle. They can integrate with MCP-based systems, knowing that another company, another tool, or even another industry can connect to that same interface without starting from scratch.

And they can do it under an open governance model that ensures no single company controls the rules of engagement.

I believe this is a meaningful shift. It means AI no longer has to be a proprietary experiment. It can become part of a production infrastructure that is reliable, transparent, and built for scale.

The WAV Group Perspective

For WAV Group, this development signals a clear direction. The conversation about AI in real estate is no longer just about features or tools. It has transitioned to be about standards, governance, and long-term architecture. Similar to what the industry has done with a standard data dictionary and transport from RESO.

The companies that take this seriously and see AI as infrastructure rather than novelty will be the ones that lead the next phase of industry transformation.

We’re already helping brokerages, MLSs, and vendors explore this shift. We are designing strategies that align with MCP, integrating agentic workflows using frameworks like goose, and helping teams write their own AGENTS.md playbooks.

If your organization is exploring how to bring AI into your ecosystem safely, effectively, and with lasting impact, now is the time to engage.

Partner with WAV Group to align your AI strategy and implementation with the standards shaping the next generation of real estate technology.

The post Agentic AI’s Next Standard and Why the Agentic AI Foundation Matters for Real Estate appeared first on WAV Group Consulting.

That’s when reality hits.

The True Cost of “Smart”

Imagine one AI agent running 1,000 requests a month. That’s roughly 20 million tokens if we average 20,000 tokens per request. Let’s assume out of the 20k requests, 15k are input, 5k output, and assume 30% of the input is cached.

Using a consumer AI model (LLM) like ChatGPT, Grok, Claude, or Gemini, that’s invisible. At enterprise scale, it’s a budget line item that can add up.

Monthly Cost Breakdown by Model (1,000 Requests)

NOTE: Costs displayed are at the time of publishing this article

Tokens Per Request Example

To put the requests-to-tokens relationship in perspective, I recently spent 10 days building a voice-first AI experience to put several large models through their paces.

My goal? Cut through the hype and see, firsthand, how quality stacks up against cost when you move beyond the demo phase. The Gemini 2.5 Flash Native Audio Dialog model, in particular, offered some eye-opening insights.

Since this was strictly a proof-of-concept, I ran everything on a free-tier account.

Shoutout to Google for offering real features and generous limits, even at zero cost.

For this article, I’m focusing on request and input tokens only (output tokens still hit your wallet if you scale up).

In just ten days, input usage topped 910,000 tokens across only 58 requests. The prompts? Nothing wild—just standard test queries. Still, that averages out to a whopping 15,700 tokens per request.

If this hadn’t been on a free plan, input alone would’ve cost me just under thirty cents. That’s pocket change for solo testing in your spare time.

But scale that up. Say, you’re running 20 sessions, 100 requests a day. At 15,700 tokens per request, you’re suddenly looking at 31.4 million tokens daily, almost 1 billion a month. At $0.50 per million tokens, input alone could set you back $471 each month.

Most AI Tasks Don’t Need a Ferrari

Let’s be blunt! Most tasks that MLSs and brokerages want to automate are routine, high-volume, and perfect for Nano or Mini models. Here at WAV Group, when we develop your AI applications, we build in optionality that enables you to associate the least expensive LLM model for the best result.

For example, when normalizing data across thousands of listing entries each day, the task is predictable and structured. An ideal fit for a low-cost model that can handle field validation and correction with speed and consistency.

When running listing audits to identify missing photos, incorrect room counts, or inconsistent property descriptions, there is no need for deep reasoning. All that is needed is fast, scalable text and image processing.

In member support Q&A systems, most questions concern office hours, login issues, or rule clarifications. A mini model can easily achieve high accuracy on those tasks using a knowledge base or fine-tuned embeddings. Deep reasoning is not required to look up a fact.

Filling out forms based on prior responses or public record lookups is another area where a simple agent can shine. The task is structured, repetitive, and also does not need advanced reasoning.

Even internal search across MLS documents, training guides, or help desk archives can be handled effectively with lightweight embedding and retrieval workflows, keeping costs down while improving access to institutional knowledge.

None of those need a GPT-5.2 model that costs nearly $90 per month per agent for just 1,000 requests. What enterprise brokers and MLSs should know is that you can save your agents lots of licensing fees by delivering AI at scale rather than each of them paying for one or more LLM products.

Reserve Premium Models for High-Stakes Work

There are moments when you do want the Ferrari.

When interpreting new or evolving regulations that impact brokerage operations, accuracy and nuance are critical. A premium model can absorb complex legal phrasing and return contextual summaries that support compliance efforts.

If you’re drafting emails, press releases, or official statements on sensitive topics, such as fair housing violations or legal disputes, a top-tier model helps strike the right tone while ensuring consistency and professionalism.

When creating polished content for executive presentations or investor updates, nuance and clarity matter more than speed. A higher-end model can improve grammar, align with tone, and provide suggestions that elevate the narrative.

Strategic generation is another high-value use case. If you’re feeding in a mix of market data, internal KPIs, and partner feedback to surface trends or recommend direction, you want a model that can reason across unstructured inputs and still deliver an actionable output.

Reserve premium models for these use cases, and deploy them only when it matters most.

What Consumer AI Gets Wrong

Consumer AI trains people to think “always use the best.” You never get throttled. You never see a bill. There’s no feedback loop.

But enterprise AI? You’ve got to think like an operator. Every model call has an impact. Every task needs to justify its cost.

Consumer AI isn’t the only game in town. You can self-host SLMs and LLM models either on-premises or in the cloud, or you can spin up GPU cycles on demand. Better yet, you can fine-tune these models to reflect your company’s tone, governance, and culture, shaping them to fit your business like a glove to bring cost efficiency in running them. Moreover, you can connect AI to useful tools that are already in your tech stack – from basic things like sending an email, setting a calendar appointment, building a presentation, to more complex activities like setting up a saved search or drafting an agreement. See CompassAI for examples.

There’s a whole world beyond plug-and-play APIs, and we’ll dig deeper into these strategies in future articles.

The Operational Playbook

If you’re serious about building AI into your operations, you need to approach it strategically.

First, architect your systems for flexibility. Don’t assume one model fits every need. Design workflows that can route tasks to different models based on complexity, urgency, and cost sensitivity.

Second, automate your cost intelligence. Set up dashboards or logging systems that show exactly how many tokens are being used, by whom, and for what types of tasks. This visibility helps you optimize spending and improve the accuracy and efficiency of your AI models.

Third, segment your tasks thoughtfully. High-volume, low-risk operations should run on cheaper models. Save the expensive models for when they’re truly needed.

And finally, think like a product manager. Each model call is not just a utility, it’s a feature with costs, risks, and returns. Evaluate it that way.

And above all, treat AI as a managed cost center. Because if you don’t, it’ll quietly eat your margin alive and profits will fly away.

If you plan to get started with AI in 2026, or you would like to roadmap your expansion of AI use in your brokerage or MLS, we are ready advisors and can either supervise or perform your development. At WAV Group, you always own your AI.

The post 💸 💸 AI Token Costs Are Invisible Until They Aren’t appeared first on WAV Group Consulting.

Introducing WAV Group’s new white paper on AI-powered pricing strategy.

The CMA has been the anchor of real estate pricing for decades. It is familiar, trusted, and indispensable. But the way it is produced today has not kept up with the complexity of the modern market. Agents still run a search, check a few boxes, export a PDF, and call it a day. Meanwhile, the MLS holds data that could transform pricing intelligence overnight.

Our new white paper, The Future CMA Will Be Powered by AI Analysis and Backed by Human Agent Oversight, is written for MLS executives, brokerage leaders, and technology innovators who are ready to rethink the CMA altogether. The paper outlines how today’s CMA falls short, why AI is the catalyst for a complete rebuild, and how the industry can construct the next generation of pricing tools from the ground up.  

DOWNLOAD HERE

Who this paper is for

This paper is designed for

• MLS leaders who want to deliver more value to subscribers and modernize their core product offerings.
• Brokerage executives who are looking to differentiate their agents with better listing intelligence and stronger pricing narratives.
• CMA vendors and proptech founders who need a roadmap for building tools that go beyond comps and into true market strategy.
• Policy and data leaders exploring how buyer intent data, mortgage rate changes, and listing engagement signals can be responsibly integrated into pricing systems.

If you are building technology, setting strategy, or guiding listing agents, this paper gives you the blueprint for what comes next. If you need help with development, fill out the form below.

What the paper delivers

The white paper lays out a practical, forward-looking roadmap for transforming the CMA from a backward-looking report into a living pricing system. It explains how:

• AI can analyze photos, MLS data, and market patterns at a scale human agents never could.
• Computer vision can quantify curb appeal, lighting, staging, and layout flow.
• Bayesian-style models can update pricing automatically as new sales close.
• Dynamic CMAs can react to mortgage rate changes in real time, something that simply does not happen today.
• MLS buyer behavior, such as saved searches and favorite listings, can become a demand-side signal that finally informs listing strategy.
• Presentation layers like Canva can turn raw intelligence into compelling narratives that help agents win trust in the living room.

The paper makes one point very clear: a CMA of the future is not a PDF.

It is a continuously updated story that reflects live economics, real buyer interest, and precise listing strategy.

Why this matters now

Agents, sellers, and MLSs all felt the impact of the Federal Reserve’s recent rate cut. Payments shifted overnight, but listing prices in the MLS barely moved. Not one CMA updated automatically, even though every listing in America should have had a pricing conversation the next morning. The gap between what the MLS knows and what the CMA shows is widening.

This paper explains how to close that gap with accessible, achievable technology that MLSs and vendors can begin developing today.

A call to industry leaders

The future CMA will not be won by the company that generates the prettiest report. It will be won by the one that builds a dynamic, data-aware pricing engine that keeps up with the market and tells a story with a listing strategy sellers can trust.

DOWNLOAD HERE

If you want to understand what that tool looks like and how to build it, this white paper is your roadmap. If you need help, then please don’t wait – reach out below!

The post Hello MLS – The future CMA is here, and it is no longer a report appeared first on WAV Group Consulting.

Real Estate Webmasters is no stranger to the competitive world of real estate technology. Serving brokerages, teams, and top-producing agents, the company has built its reputation on delivering custom solutions tuned for organizations where sales performance is measured in real dollars, not demo slides.

Meeting the sales challenge

For high-performing agents, growth depends on two distinct motions: staying connected with past clients and networks, and creating new demand. Too often, technology platforms treat these as one workflow. Real Estate Webmasters recognizes the need to separate them and is building tools that respect the reality of how elite sales teams grow.

Conversations: genai inside the crm

This month, Real Estate Webmasters introduced “Conversations,” a Generative AI solution embedded directly in their CRM. It’s designed to:

• keep past clients engaged with personalized, timely touches
• help agents respond faster to new opportunities with suggested first touches and follow-ups
• give managers insight into live conversations with summaries and coaching recommendations

By bringing AI into the CRM itself, Real Estate Webmasters avoids the distraction of standalone tools and makes AI a natural part of the sales process.

Leadership insight

In our interview, Morgan Carey, CEO of Real Estate Webmasters, explained how Conversations was designed to accelerate both the “nurture” and “acquisition”. He highlighted the role that AI can play in helping agents scale their best practices without losing authenticity or compliance.

Watch the full discussion below to hear Carey describe how Real Estate Webmasters is blending technology with sales discipline, and why he believes Conversations will become a core driver of revenue growth for leading brokerages and teams.

Connect with Real Estate Webmasters

If you’d like to explore how Generative AI can impact your organization, reach out to Morgan Carey directly. Real Estate Webmasters is eager to help brokerages and teams deploy technology that creates measurable outcomes, not just buzz.

The post Real Estate Webmasters launches “Conversations” to boost sales growth with GenAI appeared first on WAV Group Consulting.

That’s why associations need to pay close attention to the next wave of technology adoption. Generative AI, powered by graph databases (often called Graph RAG), is reshaping how professionals access critical information. And it’s happening fast.

What Graph RAG can do for agents

Graph RAG solutions allow an agent to ask plain-language questions and receive authoritative, context-aware answers instantly.

• What does state law say about seller disclosure?
• How does my REALTOR® association’s policy apply to this situation?
• Which MLS compliance rule governs this listing scenario?
• What contract clause controls if a buyer misses a deadline?

Instead of sifting through PDFs, waiting on a helpline, or misinterpreting a clause, agents can get trusted, on-demand guidance in seconds. For a professional who’s in the middle of a client negotiation, that kind of clarity isn’t a convenience, it’s mission-critical.

The gap in delivery

The irony is that REALTOR® associations already own the content that powers these answers. State statutes, association policies, MLS rules, and standard contracts are created, curated, and governed by these organizations. Yet, the technology delivering AI-driven access to that content is being built elsewhere.

Brokers, vendors, and outside platforms are moving first. They recognize the demand and are creating solutions that strengthen their own agent relationships. Associations, meanwhile, are watching from the sidelines while their authority is repackaged and distributed without their leadership.

Why this matters now

Agents increasingly question what they get for their dues. Many see associations focus based on advocacy and compliance enforcement, rather than providing daily utility. If associations fail to step into the AI space, they risk becoming even less relevant to agents’ day-to-day business.

Offering AI-powered policy and contract bots as a member benefit changes that equation. It makes the association the first stop for answers that matter in real time. It strengthens trust by delivering authoritative responses directly from the source. And it ensures that the association is present in the daily workflow of its members, not just during renewal season or political campaigns.

A new kind of member benefit

Over the years, associations have added value through lockbox systems, transaction platforms, and forms libraries. Each innovation became a benefit members couldn’t imagine working without. AI bots could be the next.

By making these solutions a member benefit, associations have the chance to:

• Provide daily, indispensable utility to every member
• Reassert their role as the authoritative source of truth
• Strengthen engagement and loyalty at a time when both are fragile
• Ensure compliance and accuracy by controlling how policies and rules are surfaced through AI

The path forward

The opportunity is immediate, but so is the risk of inaction. Associations that fail to deliver this benefit may watch brokers and vendors seize the role of trusted daily guide. The time to act is now.

WAV Group is already working with MLSs, brokerages, and technology companies to build AI frameworks that respect compliance, protect data ownership, and deliver tangible value to professionals. REALTOR® associations are uniquely positioned to do the same by turning their own policies, contracts, and regulations into the most valuable member benefit they’ve ever offered.

If your association wants to stay indispensable, the next step is clear: partner with WAV Group to build the AI bots your members will soon rely on every day.

The post AI bots could be the next must-have member benefit for REALTOR® associations appeared first on WAV Group Consulting.

I don’t like writing scare pieces. But this one? It needs to be written.

Because if your team is deploying AI agents or leveraging AI desktop tools using the Model Context Protocol (MCP) and you’re not securing them with a gateway, you’re basically leaving the doors and windows open and walking away.

So, what is MCP and why should I care?

The Model Context Protocol (MCP) is like the glue that connects AI agents to outside tools and information. It lets an AI model talk to your CRM, hit your internal APIs, or fetch files on your system.

Sounds useful, right?

It is. That’s why so many teams, from startups to massive enterprises, are adopting it. MCP makes AI agents way more capable. It turns them into doers and not just talkers.

But there’s a catch.

MCP servers require security considerations

A recent security assessment by Equixly looked at dozens of popular MCP implementations. The results weren’t promising:

• 43% had command injection flaws
• 30% allowed Server-Side Request Forgery (SSRF is basically letting attackers poke around your internal network)
• 22% exposed arbitrary file read vulnerabilities
• Only 30% of vendors even patched the issues when they were told

Worse? Some vendors claimed these risks were “theoretical” or “acceptable.” That’s like a car company saying exploding airbags are “edge cases”, and only happen when there’s an accident.

These are not theoretical. They’re real. And they’ve already caused real-world incidents.

The hacks are creative and terrifying

Let’s break down what’s happening out there:

• Prompt Injection: Attackers can sneak commands like “IGNORE ALL PREVIOUS INSTRUCTIONS” into API responses. Your AI agent happily obeys.
• SQL Injection: Old-school attack, new playground. Some MCP servers let you drop malicious SQL into prompts and exfiltrate data.
• Cross server shadowing: MCP metadata or responses change how the AI interacts with other servers.
• Server Spoofing/Tool Mimicry: MCPs trick the AI into using the wrong servers & tools.
• Authentication Bypass: Some servers don’t verify who’s calling. Others let you register rogue MCP endpoints and impersonate trusted tools.
• Tool Poisoning: A tool looks safe at install. Then one day, it updates silently and starts stealing data.
• Rug Pulls: Third-party MCP packages switch behavior after getting adopted widely—just like malicious npm packages have done for years.

This isn’t speculation. It’s already happened as detailed in security investigations from Composio and Equixly:

• One attack chain exposed Asana data via unsecured MCP endpoints
• Another let attackers run remote commands on public-facing servers
• One even granted access to private GitHub repos through a compromised MCP tool

Here’s what actually works: The MCP Gateway

Gateways act like bodyguards for your AI agents.

They sit between the AI client and the MCP server. Every request goes through the gateway. Every response does too.

The idea is simple: Centralize security. Remove trust from the server layer. Lock everything down.

Here’s how they help.

1. They handle identity properly

• Full OAuth 2.0/2.1 support
• Short-lived tokens (so even if someone grabs one, it’s useless soon)
• Role-based access control
• Integration with enterprise identity systems like Okta, Azure AD

Your AI agents don’t manage auth. The gateway does. That’s safer and way easier to manage.

2. They validate and sanitize everything

This is the magic. The gateway checks:

• Are prompts malicious?
• Is someone trying to inject SQL or shell commands?
• Are any tool descriptions poisoned?

It also strips out anything sketchy. Think of it like a metal detector for every request.

Some even use machine learning to detect suspicious prompts.

3. They audit, monitor, and alert

Every request. Every response. Logged.

You can get real-time alerts when something fishy happens. You can plug into your SIEM. You can see what tools were called, by whom, when, and how.

This isn’t optional anymore. It’s table stakes for enterprise deployment.

4. They lock down the tool supply chain

Before a tool is allowed through the gateway, it’s scanned:

• What’s the source?
• How popular is it?
• Has it ever been flagged?
• Is the repo still active?

Tools that fail checks can be blocked automatically.

If you’re not scanning tools, you’re just waiting to be breached.

So who’s building these gateways?

There are a number of gateway solutions now available, offering different levels of security, specialization, and enterprise readiness. Below are several strong options:

Enkrypt AI Secure MCP Gateway

Offers dynamic tool discovery, built-in prompt sanitization, and enterprise-grade authentication for secure MCP deployments.

• Built‑in security scans
• Dynamic tool discovery
• Works with enterprise authentication
• Performance‑optimized

Lasso Security MCP Gateway

Focuses on threat prevention with:

• Plugin architecture
• Server and tool risk scoring
• Automated blocking of high‑risk components

WAV Group Gateway Template (Real Estate Focus)

WAV Group offers a Gateway Template designed for real estate brokerages and MLSs. Key features:

• Prompt sanitization tailored for real estate contexts
• Guardrails for private client/buyer/seller data
• Role‑Based Access Control (RBAC) at agent/user levels
• Audit logging specific to real estate workflows
• MLS API integration controls and PII masking for real estate data
• Designed as a template clients can adopt to deploy secure, compliant AI agents in real estate environments

Obot MCP Gateway

Obot is an open‑source gateway focused on enterprise requirements. Some of the features:

• Admin control plane: IT can onboard MCP servers, define access policies, manage users/groups, monitor usage. 
• Catalog / discovery: A searchable directory of approved MCP servers, documentation, trust/reputation information. 
• Proxying & hosting: Support for both local and remote MCP servers; ability to proxy third‑party ones with audit and routing control. 
• Access control + logging: Role‑based access, enterprise auth integration (Okta etc.), audit logs for MCP‑client/server interactions. 

Kong Konnect / Kong AI Gateway

Kong is more known as an API gateway, but it’s also building out MCP support and gateway‑style features. Key capabilities:

• Kong Konnect MCP Server: Enables MCP clients (e.g. Claude) to query APIs, configuration, analytics via Kong’s control plane. 
• Securing & governing MCP traffic: Kong’s AI Gateway offers plugins and policies for authentication (OIDC / Key Auth), rate limiting, prompt filtering (guardrails) etc. 
• Observability: Metrics, logging, tracing for MCP traffic. 

What should your team do right now?

If you’re deploying MCP servers, or building on top of them, here’s a basic security checklist:

• Set up a gateway (before anything goes live)

This is non-negotiable. Even for internal tools.

• Use proper auth

Hook into OAuth. Integrate with your identity provider. Don’t hand-roll this.

• Validate inputs and outputs

Use JSON schemas. Sanitize tool responses. Strip out embedded commands.

• Lock down your network

Log everything. Store audit trails. Send alerts when strange stuff happens.

• Don’t trust tools blindly.

Scan them. Review their source. Watch for updates. Use a reputation system.

The future isn’t secure by default

MCP is a powerful idea. But it’s dangerously naive out of the box and can expose your most valuable asset, your data.

Vendors are moving fast. Too fast. And when 43% of servers have command injection flaws, you don’t get to say “well, we trust our stack.”

You lock it down. You build defensively. You audit, scan, and restrict.

This isn’t optional if you’re serious about deploying AI in production.

And finally: stop hoping and start securing

Hope is not a security strategy. “No one would ever target us” is how breaches happen. “It’s just a proof of concept” becomes a Common Vulnerabilities and Events (CVE).

The MCP ecosystem is still young. That means you get to choose your architecture now before someone else chooses it for you via an incident report.

So choose wisely.

Start with a gateway.

The post The Hidden Risk in MCP Servers That Could Expose Your Business appeared first on WAV Group Consulting.

The team specializes in full-stack software development, data migrations and integrations, and AI application development. We have helped businesses scale effectively and innovate faster.

Software Development Consulting Experts in Full-Stack, Data Integration, and AI Automation

This team has a broad and versatile skill set, enabling them to tackle complex software development consulting projects seamlessly. They are proficient in every phase of development—from front-end design to back-end engineering with API development.

Throughout our collaboration, I’ve managed multiple projects that demonstrated their remarkable flexibility, reliability, and technical expertise.

Full-Stack Software Development Consulting

The team excels in modern front-end technologies like React, Angular, and Vue, consistently delivering user-friendly, engaging applications. Their backend expertise covers Node.js, Python, and Next.js, making them adaptable to various technology stacks.

Additionally, they are experienced in deploying and scaling applications efficiently on cloud platforms such as AWS, Azure, and Google Cloud.

Data Migration & Integration Consulting

Data integration and migration are core competencies of the team. They’ve consistently handled secure, smooth, and efficient migrations from legacy systems to modern architectures. Whether it’s CRM integrations, accounting systems, or complex database migrations, the team expertly manages large datasets, ensuring minimal disruption and maintaining maximum data integrity. I’ve observed them successfully simplify complex migration processes, significantly easing client concerns.

AI Application Development Consulting

AI solutions have become essential to competitive businesses. The team is deeply skilled in the latest AI technologies, including machine learning, deep learning, natural language processing (NLP), and generative AI. We have delivered several AI-driven projects, such as intelligent chatbots, predictive analytics systems, and advanced decision-support tools. We are leveraging AI to improve our workflow in order to delivery products and solutions very quickly.

Our AI expertise has significantly enhanced business efficiency, reduced costs, and improved strategic information accuracy, enabling our clients to make solid business decisions.

Why Choose Our Software Development Consulting Team?

Having directly managed projects with these engineers, I’m confident we can deliver technical excellence and seamless integration into your team or as a standalone team for your project. We emphasize clear communication, transparency, and practical solutions, ensuring fast and efficient deployments without unnecessary complexity.

• Business-focused consulting that aligns technical solutions closely with your business goals, ensuring measurable impact quickly.
• Reliable & timely delivery in meeting tight deadlines with high-quality results.
• Flexible engagement models offering project-based and dedicated team engagements to accommodate evolving project needs.

Would you be ready to Start Your Software Development Project?

If your organization is planning software development projects, data migrations, integrations, or AI implementations, our consulting team is prepared to assist you. After nearly two years of successful collaboration, we are confident in our ability to deliver effective solutions for your business.

Get in touch today! David Gumpper. Let’s talk about your project’s needs and explore how our software development consulting team can help you achieve your goals.

The post Software Development Consulting Team Now Available appeared first on WAV Group Consulting.