Good technology projects might never leave the launchpad. 

Think of a tech initiative like a rocket launch: 

• With too little fuel, you never reach orbit. 

• With too much weight, you never get off the ground. 

Organizations shelve strong ideas for AI, workflow modernization, and data initiatives not because the technology doesn’t work, but because inaccurate estimates fail to convince leadership. Competitors that scope more realistically move ahead, while incorrectly scoped projects stay grounded. 

The ripple effects are significant: 

• Missed innovation windows. If a project stalls, competitors will move ahead and deploy automation, streamline workflows, or launch AI-driven services that capture market share.
• Wasted organizational energy. Teams that invest time drafting proposals that fail in committee create frustration and disengagement.
• Leadership skepticism. Boards that see repeated cost overruns or inflated requests grow reluctant to fund future initiatives, even if the project is strong. 

The result? Organizations remain stuck with legacy systems while competitors push forward with modernized, AI-enabled operations. 

Test your numbers and scope your project with experts 

The good news? Teams can avoid estimation failures by validating their assumptions before budget requests hit the review board. 

Here’s how hiring a third-party expert adds value: 

• Technical validation. Architects spot hidden costs in data pipelines, integration complexity, or system performance before they have any potential to surface mid-project. 
• Right-sizing scope. Instead of budgeting for an all-or-nothing build, break projects into testable phases. Smaller asks earn faster approvals and create steady progress that builds trust with stakeholders.
• Credible roadmaps. Expert scoping builds confidence. Boards trust numbers that account for risks, dependencies, and measurable milestones. 

Hiring an expert is especially important for AI projects. Boards want transformation, but many underestimate the costs for preparing data, managing inference at scale, or integrating AI into existing workflows. Without realistic estimates, these proposals can get labeled “too risky” and cut before they begin. An experienced AI agency will know how to scope projects with both the technical and business realities in mind, transforming a technology project wish list into an executable plan. 

How to strengthen your technology requests 

If you’re preparing technology proposals for 2026, take these steps to keep your requests strong:

Tie projects to measurable outcomes: Don’t just ask for funds to “implement AI” or “modernize systems.” Frame your request around a business KPI. Define before-and-after metrics and give leadership something tangible to measure.

Start small, prove value: Instead of asking for $2M to “transform the entire workflow,” budget $300k to modernize a single high-friction process. Identify a decision point in your process and build the case for improving it. Quick wins show ROI early and make the case for expansion. Iterative, right-sized delivery reaches value faster. McKinsey reports that high-performing IT organizations complete a medium-sized change from idea to production in 2-4 months, while less-advanced peers take up to a year. This is why phasing scope beats big bang bets every time. 

Validate with experts: Bring in architects or engineers to check your scope and costs. They’ll catch integration issues, data challenges, and compliance gaps that could otherwise tank your project. Hiring a third-party expert can help prevent rework, cost overruns, and potential rejection during review. 

Start scoping out your project today  

Budget season is more than paperwork; it’s where technology strategy gets tested. The difference between rejection and approval comes down to how well you scope, size, and justify your technology project. 

 At Rōnin Consulting, we help leaders validate their technology initiatives before they hit the review stage. From AI pilots to modernization, we will break projects into realistic phases, uncover hidden costs, and produce roadmaps that your board can trust. 

Great technology ideas shouldn’t stall in planning; they should get funded, built, and deliver results. 

At the Nashville Technology Council (NTC) Innovation Summit event this month, we met Santiago Campoverde, a graduate student at Middle Tennessee State University (MTSU). What struck us wasn’t just his excitement for attending the event, but his story: a student actively bridging the worlds of software development and business, eager to learn, connect, and contribute to the Nashville technology community. 

As a company founded by two MTSU graduates, staying connected to the next generation of talent matters to us. After the NTC event, we reached out to Santiago to get a student’s perspective on what it’s really like out there in Nashville right now: balancing school, career goals, and a field dominated by AI. 

Hello Santiago! Tell us a little about yourself. 

Santiago: My name is Santiago Campoverde, and I’m 24 years old. I’m currently pursuing a Master’s degree in Information Systems at MTSU. I came here after completing my undergraduate degree in Software Engineering at Iowa State. I’ve always loved technology, but one thing I realized while studying was that technical skills alone don’t always translate well into the business world. That’s why I wanted to expand my perspective, learning not just how to code, but how to connect technical work to business needs. 

What brought you from Iowa to Nashville? 

Santiago: Iowa is a great state, it’s beautiful, but I wanted to be closer to a city and more opportunities. When I discovered MTSU’s graduate program and saw that it offered an IT assistantship, I knew it was the right choice. It checked all the boxes and gave me a chance to continue learning while also gaining hands-on experience. Not to mention that the Nashville technology community is very vibrant! 

You went from an undergraduate in software engineering to a graduate degree in information systems—what sparked that shift? 

Santiago: During my undergraduate studies, I led a hackathon club and began to notice a consistent pattern: many engineers were very technical but struggled when it came to communicating with businesses or stakeholders. That experience opened my eyes to the fact that it isn’t just about the code, it’s about understanding the people who will use it and the organizations it will impact. My senior project drove this home even further, as I saw firsthand how mismatched expectations between developers and stakeholders could stall progress. That’s when I realized I wanted to be that bridge, someone who understands the technical side but can also translate it into business solutions. 

Where do you see yourself in the future? 

Santiago: I believe that a project management or a business analyst role would really fit what I want to pursue. I still love problem-solving and coding, but I see the gap between business needs and technical execution as something that must be solved. My goal is to be flexible and lean into the technical side when needed, while also translating that into solutions that make sense for businesses. 

What brought you to the Nashville Technology Council event? 

Santiago: My graduate assistantship with Carlos Coronel at the MTSU College of Business actually opened the door. He encouraged me to go to the event. And, since I’m new to Nashville, it was an excellent opportunity to meet companies, make connections, and learn about the Nashville technology scene. It has really helped me feel a part of the community. 

What stood out at the Nashville Innovation Summit? 

Santiago: The energy was amazing. I met a lot of great people that I wouldn’t have met if I hadn’t gone. It was both fun and motivating. I saw how much was happening in Nashville’s software development, artificial intelligence, and technology sectors, and how open people are to connecting. 

Let’s discuss AI, the current hot topic in the industry. What’s your take? 

Santiago: At first, I was hesitant. I didn’t want to depend on AI and lose sight of my core skills. However, over time, I’ve come to view it as a tool, something that, when used wisely, can enhance a developer’s abilities. At MTSU, we are encouraged to explore AI technologies like Copilot, Perplexity, and Gemini. For me, though, the key is balance: use it, but don’t let it replace your own understanding. I’m hoping to explore AI further, but I’m still developing my core skills. 

You also walked away as our guitar giveaway winner! What’s the story there? 

Santiago: Yeah, that was awesome. I was entered into your drawing, and I was the first ticket you pulled out! It was a complete shock, but I was so happy to win! My dad is a big guitar player, and I used to play too, but I had a wrist injury that made me stop for a while. Winning that guitar felt like a sign to pick it back up again. It’s been great to get back into playing some of the songs I used to love. 

1. Should you use a large language model (LLM) or traditional machine learning (ML)?
2. How do you measure ROI on a generative AI project?
3. Should you be worried about data security and compliance with AI?
4. How fast can you deploy AI and see value?
5. How do you estimate token costs for AI projects?
6. How do you choose the right AI model for your project?
7. Do you have to use your cloud vendor’s model for your AI project?
8. Which cloud vendor is right for AI integration?
9. Do you need a RAG pipeline (Retrieval-Augmented Generation)?

Answers to these questions are not just technical choices; they’re strategic moves. Get them right, and your AI initiative delivers real, lasting impact. Get them wrong, and your momentum could stall.

In the sections ahead, we break down these questions and guide you from curiosity to confidence in the next phase of your AI implementation.

Should you use a large language model (LLM) or traditional machine learning (ML)?

This is one of the most critical early decisions in any AI journey. It affects how you approach your data, what problems you can solve, and how much your solution will cost. In short, we like to break it up this way:

• Use an LLM when your project requires natural language understanding, summarization, classification, content generation, or knowledge extraction.
• Use traditional ML when working with structured data, like spreadsheets or databases, and need predictive models (e.g., sales forecasting or churn prediction).

Pro tip: Many successful enterprise solutions combine both. They use LLMs to handle unstructured user input and ML to handle structured, backend decision-making.

How do you measure ROI on a generative AI project?

You can’t manage what you don’t measure. And without proving the business impact, AI initiatives quickly lose support. However, ROI will take time, so it’s best to take a bird’s-eye view of your entire project and (depending on how you implement your AI) track results across these three pillars:

• Efficiency gains: Are you saving time and/or reducing labor costs through automation?
• Revenue impact: Is AI driving more conversions, increasing upsell potential, or enabling new offerings?
• Risk reduction: Has AI helped reduce human error, improve compliance, or respond to customers faster?

Avoid hype metrics that say little about real-world value, such as flashy numbers like model size or benchmark accuracy. Use real outcomes to gain buy-in and to justify scaling. Track the metrics related to the problem you want AI to solve. Measuring the issue first gives you a solid foundation to build from.

Should you be worried about data security and compliance with AI?

Yes, and you should plan for it early. Trust drives adoption, and that trust depends on how well you protect your data.

We recommend:

• Preventing training misuse: Make sure your data isn’t being used to train third-party models. Enterprise agreements can help with this.
• Securing your infrastructure: Use private endpoints, encryption, audit logging, and strict access control.
• Consider on-premises: For highly sensitive environments, deploying models on-premises offers full control. This is where understanding private vs public AI becomes especially important.

Bottom line: Own your data and protect your systems. Prioritizing security early can also help you avoid unexpected compliance headaches down the line.

How do you estimate token costs for AI projects?

Generative AI models charge based on tokens the model reads and generates. If you’re not careful, costs can escalate fast.

Watch for:

• Model size: Model size can vary depending on the project. For reasoning models, there’s typically a “reasoning effort” setting, like low, medium, or high, which affects how many tokens the model uses. Higher effort means more tokens and less predictability in output length. We usually control the “max completion tokens” setting to manage this. This sets an upper limit and helps prevent the model from going off the rails, but it can also cut off longer responses. So, the right choice depends on the project’s goals and prompt length.
• Prompt length: Bigger queries and responses = more tokens = higher costs.
• Usage volume: Frequent usage across teams can multiply your monthly bill.

To pick the right model and understand your likely costs, test early. Run simulations during a pilot phase to forecast and manage budget expectations.

Pro tip: Work with an agency that’s been through this before. A seasoned consulting partner can help you choose the right model, steer you away from wasteful usage, and guide you toward cost-effective options; including newer models or fine-tuning strategies that fit your budget.

How do you choose the right AI model for your project?

There are so many models out there, the choices can be overwhelming. The best approach is to define what success looks like for your project, then work with your agency to select a model that fits your business needs and technical requirements. Your goals, constraints, and data will shape the right model. A good place to start is to evaluate options based on:

• Data sensitivity: If you’re in a regulated industry (e.g., healthcare, finance), consider open-source or self-hosted models for better control and compliance.
• Performance needs: Latency, accuracy, and uptime can vary widely between models. Choose what aligns with your user experience goals.
• Budget: Larger or proprietary models (like GPT-4) often deliver better performance but come with a higher price tag.
• Adaptability: Some models are easier to fine-tune or integrate into your workflow, especially if your use case is highly specialized.
• Vendor lock-in: Flexibility matters. Choose models or platforms that won’t box you in long-term if priorities shift.

There’s no one-size-fits-all model—just the one that aligns best with your priorities.

Do you have to use your cloud vendor’s model for your AI project?

No, but it’s often convenient. Cloud-native models (like Azure OpenAI or AWS Bedrock) offer seamless billing, scaling, and security. However, you can also:

• Bring your own model
• Deploy open-source models
• Use a hybrid approach to optimize performance and cost

You’re not locked in. If you’re using AWS and want to go all-in on ChatGPT, that’s totally valid. But in many cases, it’s easier (and more cost-effective) to stick with the ecosystem you’re already in. Switching later isn’t impossible, it’s just harder and more expensive than planning ahead.

Which cloud vendor is right for AI integration?

The cloud platform you choose can shape everything—from model access to scalability and governance.

Here’s a simplified comparison:

• Azure: Great for companies already using Microsoft. It offers direct access to OpenAI models and native integration with Office 365 tools.
• AWS: Offers maximum flexibility and a wide range of model options through Bedrock.
• Google Cloud: Known for AI research leadership and Vertex AI; ideal for teams already invested in Google’s ecosystem.

Match the vendor to your internal tooling and team preferences, and you will save time and money down the road.

Do you need a RAG pipeline (Retrieval-Augmented Generation)?

If your AI needs to provide up-to-date, personalized, or domain-specific information, then yes, a RAG pipeline is essential. RAG enables your LLM to “look up” relevant data at query time, without retraining the entire model. This approach is:

• More accurate for enterprise-specific use cases
• Lower cost than frequent fine-tuning
• Easier to maintain with real-time updates

As your knowledge base grows, RAG becomes not just helpful, but foundational. Don’t skip this necessary tool; incorporate it from day one if you can.

At Rōnin Consulting, we specialize in helping clients with large, complex codebases. One of our biggest challenges is efficiently onboarding these clients, basically getting our engineers up to speed on intricate systems so we can deliver value quickly.

That’s where Claude Code shines, especially with its “Codebase Q&A” feature, which has become one of my favorite tools for speeding up this process.

Become indispensable infrastructure (tool/API provider).

Your SaaS should ensure it offers robust APIs or other machine interfaces because agents will be operating on behalf of humans and need programmatic access​. A product that is not easily accessible to an autonomous agent will be skipped. Forward-thinking SaaS companies are already moving this way. For example, Stripe maintains a world-class API for humans and machines alike. They also built an infrastructure for AI consumption alongside their human-facing app to stay competitive.  

This dual approach – one product for human users and one for agent use – might become the standard. SaaS firms should evaluate how to expose every significant feature via API or plugins, even those that historically required a GUI. By positioning themselves as the best “tool” in a particular domain, a SaaS can ensure agents favor it for tasks, preserving usage. 

Embracing the infrastructure/tool role means that a SaaS must let go of the notion that users must see your interface. Some vendors fear losing customer engagement if automation replaces their UI. But trying to trap users in your UI is a mistake. 

In practice, this might also involve offering new integration formats (e.g., being part of popular agent frameworks, providing SDKs, supporting open agent standards) so your SaaS becomes a default building block in AI-driven workflows. 

Embed intelligence or provide it.

SaaS companies can also move into the intelligence layer by deeply infusing AI into their products. We can already see instances of this with the recent wave of built-in assistants (e.g., Salesforce’s Einstein GPT and Adobe’s Firefly in Photoshop).   

Offering an AI copilot inside your SaaS can serve two purposes: it makes your product AI-enhanced for end-users and trains the company to operate an LLM in its domain. Over time, a SaaS provider could develop proprietary models or fine-tuned AI that become its own competitive advantage.  

For instance, a SaaS with years of specialized data might fine-tune an AI that outperforms general models for specific tasks. This could effectively become a niche “intelligence” provider in its domain.  

Moreover, AI-first design is key. Rather than bolting AI on as an afterthought, rethink your application as if an AI is a primary user. Ask yourself these questions: 

• How would you redesign workflows so an agent can easily navigate them? 
• Are there internal optimizations or data pipelines you can expose to AI? 
• Can you structure your application to seamlessly collaborate with AI, allowing it to make proactive decisions and enhance user interactions? 

SaaS teams should refactor rigid logic into AI-accessible modules. This might mean transitioning some functionality from code-based rules into model-driven policies that an agent can adjust. As a venture technologist advised, “Companies must evolve from traditional architectures to AI-first platforms, enabling agents to interact seamlessly with their tools.”​ 

Own a piece of the control layer. 

Another strategy is to build the agent or orchestrator for your domain. Suppose your company has deep domain knowledge (say, project management or marketing operations). In that case, you might create an AI agent that orchestrates tasks in that domain, essentially offering “X- Automation-as-a-Service” powered by AI.  

This move is risky but potentially disruptive: it means moving from being one SaaS tool among many to being the brain that coordinates multiple tools (including possibly your competitors’ tools!).  

For example, a SaaS project management company could release an AI agent that takes high-level project goals and automatically uses Jira, Confluence, Calendar, Slack, etc., to execute the plan. Doing so, you reposition from a tool provider to a workflow orchestrator for that vertical. 

Incumbents with broad product suites are exceptionally well-placed here. They can integrate an agent across their ecosystem (e.g., Microsoft’s 365 Copilot spans Office apps). However, startups are also attempting this in niches, essentially launching “agent-native applications” that directly compete with legacy SaaS by abstracting them.   

This move represents a significant shift, transforming the traditional service-as-software model—where software provides a service—into something entirely new. However, not every SaaS will pursue this, but it’s worth chasing if your core value could be delivered as an autonomous agent that works on the customer’s behalf. Even if you don’t build the agent brain from scratch, ensuring your product can plug into popular agents as a trusted executor will be necessary. 

Maintain dual experiences: human and agent. 

In the near future, leading SaaS companies must offer two seamless experiences: a refined user interface for humans and a machine-accessible interface for AI agents. Just as web apps had to evolve with APIs for mobile integration, AI now demands a similar shift. As seen with Stripe’s AI Agent model, industries must recognize and adapt to this emerging duality. 

This shift might mean SaaS companies investing in things like AI-specific documentation, sandbox environments for agents, or even a “virtual assistant mode” of their product. While this is operationally complex, it buys time to serve current customers while preparing for the agent-dominated future. It also keeps your SaaS in the loop regardless of whether the end-user is a person or an AI. 

Opportunities for incumbents and startups 

Both established SaaS companies and new startups have opportunities in this agentic future, though their playbooks differ. 

For incumbents  

Incumbent SaaS firms have assets to leverage – data, customer trust, and domain experience. These can translate into durable positions if used wisely. One significant advantage is the proprietary data context. An AI agent finely tuned to a company’s unique dataset and workflows delivers exclusive value that competitors can’t easily replicate or monetize.​ 

An enterprise SaaS with years of accumulated domain knowledge can build an agent that performs in ways a generic tool cannot. Incumbents should double down on their data moats. For example, a CRM company can train AI models on its own aggregated (and privacy-compliant) sales interactions to offer insights that no generic CRM API ever could. They can also provide enterprise-grade assurances such as security, compliance, and reliability for AI integrations, which many CIOs will demand. 

Incumbents can turn this threat into an opportunity by introducing their own agent platforms. We have already seen early moves here: Salesforce with its AI Cloud and Einstein agents, Microsoft with its Copilots, etc. An incumbent could offer an agent that prefers its own suite of tools, creating a ripple effect across their products.  

They also have existing distribution: they can bundle AI capabilities into their plans, upsell AI features, and educate their large customer bases on using these new tools. Far from being destroyed by AI, an incumbent who adapts can strengthen customer lock-in by becoming the orchestrator of how work gets done on their platform. 

However, there is a note of caution here. Incumbents will face an innovator’s dilemma. Today, their revenue often comes from seat licenses and the human usage of their apps. Shifting to agent usage (potentially fewer human logins) might upend their monetization. To account for this, they must navigate pricing models for AI usage.  

Despite these challenges, the cost of inaction is higher, and applications that don’t adapt might face declining usage – regardless. Like the companies that failed to go mobile, those who ignore the agent trend risk becoming the next cautionary tale. 

For startups  

The AI-agent wave is a classic platform shift for new entrants that levels the playing field. Startups are not burdened by legacy UI or business models – they can build an AI-native from day one. One clear path for these startups is identifying niches or vertical workflows that big SaaS companies poorly serve and creating AI agents to automate them. 

For instance, a startup could build an agent specializing in real estate lease management or biotech research assistance – domains where incumbents are slow to adapt. By delivering tangible results (time saved, higher output) rather than just software, these new businesses can win customers who care more about outcomes than brand names. 

Startups also have the chance to build a new agent ecosystem. Every new technology wave creates demand for supporting tools. We can anticipate needs like agent monitoring and observability, security layers, and interoperability standards. Just as past SaaS eras gave rise to monitoring tools and integration platforms, this agent era will need its tooling. 

Finally, startups can aggressively align with the new value chain. They can skip building full-stack apps and instead focus on being the best at one layer. They can also pursue creative business models – for instance, usage-based pricing for tasks completed or success-based fees – essentially selling results rather than software seats. This aligns well with how an agent delivers value. 

Both incumbents and startups should recognize that this isn’t a zero-sum game of AI agents versus SaaS. It’s about those who leverage the change versus those who resist it. The pie will continue to grow with new capabilities, but slices will be redistributed. In many cases, partnerships between incumbents and startups (e.g., an old-guard company adopting a startup’s agent framework) could accelerate adaptation on both sides. 

6 practical steps for SaaS companies 

Adapting to an AI-agent-driven future can feel abstract, but here are practical steps these companies should take now to position for this future:

Expose and enhance your APIs 

Make sure every key function of your product is accessible via API (or other machine interface) and prioritize API robustness and documentation. AI agents are going to drive demand for APIs through the roof. Audit your API coverage. If features are only available through the UI, ensure they are accessible via API as well. Consider joining integration hubs or agent marketplaces to increase your visibility to AI developers. Evaluate how effectively agents interact with your API.  

Identify potential obstacles such as rate limits, authentication challenges, or output formats hindering machine consumption. Optimize these aspects to ensure a seamless, agent-friendly experience. 

Develop an agent integration strategy 

Decide how your product will plug into the agent ecosystem. This could mean building a plugin for popular AI platforms (for example, a ChatGPT plugin that interfaces with your SaaS) or offering pre-built connectors for automation tools. The goal is to reduce friction for any AI agent using your service. Begin to treat agents as a new class of customers and court them by making integration easy. Some companies are creating agent SDKs or libraries to simplify how external developers can incorporate their SaaS functionality into agent workflows. 

Launch AI copilots & assistants  

Integrate an AI assistant within your UI to augment your human users. This serves a dual purpose: it differentiates your product today and builds your internal AI competencies. These copilots can also handle multi-step tasks internally, functioning as mini-agents within your app. By doing this, you keep your UI relevant (users enjoy new AI-driven features) while ensuring that if a user’s AI agent is using your app, it can potentially interface with your AI assistant (agent-to-agent communication).  

Adapt pricing and metrics 

Begin to shift how you measure and charge for success. If historically, you charged per seat or per active user, you might consider usage-based pricing to capture value from agent usage. Monitor metrics like a number of API calls by agent systems, tasks completed via automation, etc. This will help you understand adoption in the new model.

It also signals to customers that you’re aligned with their AI automation goals (e.g., offering an option to pay for outcomes or usage, not just logins). So, adapt your business model to incentivize and monetize heavy usage, regardless of whether it comes from people or AI. 

Educate and co-create with customers 

Many customers (especially enterprises) will be cautious about AI agents. SaaS providers should proactively help them integrate AI workflows with your product. This might involve publishing guides or best practices for using AI agents with your SaaS. 

This could also mean working closely with pilot customers to build successful agent integrations – a consultative approach to ensure your SaaS fits into their AI-driven processes. By doing this, you make your product stickier and learn real-world usage patterns to refine your offerings. You should also consider partnering with forward-thinking clients to create case studies. E.g., “X Company used our API and an AI agent to automate 50% of their workload – and here’s how.” Such stories will both validate your relevance and provide feedback for improvement. 

Invest in agent-era capabilities 

Finally, look internally at what new capabilities and talent you need. This could mean hiring ML engineers or prompt engineers to improve how your product works with AI. It might involve beefing up your infrastructure to handle an onslaught of API calls from agents working 24/7.  

Consider building monitoring tools that track AI usage specifically – e.g., flag anomalous agent behavior or errors when an agent interacts so you can troubleshoot. Ensure your security model covers scenarios like an AI agent with API keys (you may need more granular permission scopes or rate controls to prevent mistakes at machine speed). Gear up your tech stack for “always-on” machine clients that will be there alongside human users. This groundwork will pay off as agent usage scales. 

Embrace the change, don’t fight it 

The rise of AI agents represents a fundamental shift in the software landscape, but it’s not a death knell for SaaS – it’s a call to evolve. As SaaS once disrupted on-premises software, SaaS companies must now reinvent themselves for the agentic age. Those who seize this moment will find new growth opportunities, whether by powering the brains of AI workflows or by automating outcomes for customers in unprecedented ways. Those that ignore it, clinging to old UX-centric models, risk becoming obsolete as AI-driven workflows route around them. 

The urgency is real – but so is the opportunity. By rethinking strategies and repositioning within the AI agent ecosystem, SaaS businesses can ensure they remain relevant and essential in the future of software. 

Adaptation is the only path forward. The agent era will reward companies that provide value in whatever form – UI, API, or AI-driven service – and punish those that rigidly stick to yesterday’s playbook. Embrace the coming changes with an open mind and a proactive plan, and you can turn disruption into a new chapter of growth for your SaaS business.​ 

Looking Forward 

Last year marked our first Hackathon as a company, and we entered it without knowing what to expect. Our only certainty was our focus on leveraging technology to solve real-world healthcare challenges.  

This year, we’re more prepared and excited to work on our solution, which is more granular than the 2023 Hackathon. This year, the task is about solving problems in the healthcare industry facing nursing professionals by leveraging AI.  Nurses are under immense pressure due to workforce shortages, increasing patient care demands, and growing administrative responsibilities.  

“Last year, we created a solution that helped streamline healthcare processes,” says Chris Bybee, CDO of Rōnin Consulting, “And this year, we’re eager to develop something that will have a meaningful impact on the challenges nurses face daily.” 

We believe technology—particularly AI—can play a vital role in alleviating these burdens and supporting those on the frontlines of healthcare. 

While we can’t share our solution yet, our team is energized by the possibilities ahead. We’re committed to designing something that will push our technical limits and provide lasting value to nurses within the healthcare community.  

As we start another hackathon, we look forward to another few weeks of collaborating with industry leaders, fellow developers, and healthcare professionals to create something truly impactful. 

If you want to learn more about our AI services and how we have implemented this technology into our business, contact us today! 

How do businesses keep up with emerging trends and technologies, particularly AI advancements?

To help me answer this question, I sat down with Enterprise Architect and co-founder of Rōnin Consulting, Ryan Kettrey. Kettrey is no stranger to emerging technologies. As a long-time software developer and perpetual student–he had plenty to say about the future of AI, his involvement within the field, and how building out a new AI offering, SamurAI^TM, feels like a natural progression of Ronin’s offerings.  

Julie Simpson  

I’ve heard you decided to pursue a post-graduate education that delves deeper into AI. What made you decide to choose this path?  

Ryan Kettrey    

Every software developer needs to stay current. If you’re in this field, you’re constantly adapting to new languages and innovative methods. While software developers learn the fundamentals in school and hone skills through on-the-job training, what happens when new technologies emerge and the industry undergoes significant shifts like what we’re experiencing now? For me, the explosion of AI marks a crucial moment where learning must intensify to meet the demand for innovation and progress. 

Julie Simpson 

Why is the emergence of AI different from other technologies? 

Ryan Kettrey    

Artificial intelligence and machine learning have existed for a long time, but the explosion of AI advancements and the demand for AI solutions in the software industry has been astounding. AI in the mainstream has expert developers up to C-level executives asking us, “Hey, what can our business do with AI, and can you help?”  

As a developer with decades of experience who was comfortable guiding people on software development practices and projects, I quickly realized that I wanted to be in a better position to guide this part of custom development. I could have learned AI through certifications and hitting the books in my spare time, but I needed something that had real-world projects in it and something more structured with deadlines.  

Julie Simpson 

Was there a specific aspect of AI that you wanted to learn that would help you be in a better position to guide AI custom development projects for Rōnin? 

Ryan Kettrey    

That’s a good question, and yes, there was a specific part of AI that I wanted to learn. I wanted to understand the upfront process of what data science is. I wanted to know how to perform data analysis to understand the variable relationships, execute all the math and statistics, and know how to build ML models from scratch.  

My counterpart, with whom I founded Rōnin, Byron McClain, has been focused on selecting prebuilt machine learning models, fine-tuning them, and then putting them into action. He is a super-talented guy who was able to adopt an understanding of AI technology.  

I knew that whatever I was going to learn had to complement Byron’s areas of focus. If he had his part down with using pre-built models and I had mine down with the data and statistics to build them, we would have the whole AI life cycle covered and could easily upskill our team on AI and offer excellent AI solutions to clients.  

Julie Simpson 

Why would you need to learn how to build these models if we can choose to use prebuilt ones?  

Ryan Kettrey  

For common problems like text summarization, a prebuilt machine learning model that we can fine-tune is an approach we will use. However, some clients have specific problems that don’t fit into a prebuilt model. I want us to be able to handle those as well.  At Rōnin, we want to provide our clients with every part of the AI lifecycle, which means we need to be good at both exploratory data analysis and model building.  

The explosion of AI marks a crucial moment where learning must intensify to meet the demand for innovation and progress. -Ryan Kettrey

Julie Simpson 

With two of the founding owners digging into artificial intelligence – how do you foresee this impacting Rōnin? 

Ryan Kettrey 

A thing about Rōnin that is so ingrained in its DNA is that we are all about solving problems. I feel like this is something that truly sets us apart from other consulting companies.  

Our folks must be jacks-of-all-trades. If we are working on a solution and run into a completely unrelated issue that needs solving, maybe even in a completely different technology, we step up and get that done, too.  

We’re problem solvers, and AI is another problem we must solve. So, for us, we could have clients come to us and say, “Hey, so I have this system; I’d like it to be able to make this prediction or have it offer summarization with AI,” and we are positioning ourselves to be able to say yes – we can do that for you, and more.  

 The “and more” part is a standalone AI offering we have been working on called SamurAI ^TM, which we plan to roll out to clients in a few months. 

Julie Simpson 

Talk about burying the headline! Can you talk more about SamurAI^TM? 

Ryan Kettrey    

Yeah, sure. In the industry right now, there are a lot of AI platforms available, but you still need to know what you’re doing. It is no easy feat to implement custom AI within your business. The process is complicated and tends to involve considerable coding, model development, and deployment.  

So, not only do you have to understand the data you want to work with, but you also need to find and choose a compatible AI platform. This also doesn’t include the associated costs, as running models is extremely expensive. To simplify much of this process, we are building our own AI offering: SamurAI^TM.

SamurAI^TM will enable companies to deploy private models tailored for their needs and integrate them at scale into their enterprise, and provide them the auditing, monitoring, and administration tools they need to manage their AI resources. 

At the end of the day, building SamurAI^TM fits with the mission we follow at Rōnin: solving business problems, consulting, and doing it using computer science and software. This new AI offering is a natural extension of that mission. It’s a new technology that we are excited to build, launch, and implement so that we can continue to advise clients and solve their problems with technology and excellent custom software. It’s a win-win.  

To learn more about our consulting services, or to speak to Ryan about an AI project, contact us today.