Building Innovation Infrastructure: A Conversation with San Jose's Chief Innovation Officer

How America's tenth-largest city is navigating the tension between Silicon Valley expectations and public sector realities. Stephen Caines is the Chief Innovation Officer and Budget Director for the city of San Jose, California.

Stephen Caines holds a unique position in American local government. As both Chief Innovation Officer and Budget Director for San Jose, California, he sits at the rare intersection where bold technological ideas meet the hard constraints of municipal finance. In a city of nearly one million residents that shares a backyard with Nvidia, Adobe, and Zoom, the pressure to innovate isn't just internal, it's baked into constituent expectations.

But innovation in city government looks fundamentally different than it does three blocks away in a venture-backed startup. It moves slower, must serve everyone equally, and lives or dies by different metrics entirely. In this conversation, Caines shares how San Jose approaches this challenge, what founders get wrong about selling to cities, and where he sees the future of civic innovation heading.

Big Picture: Public Innovation and Governance

You sit at the intersection of innovation and fiscal responsibility. How do you balance bold new ideas with the realities of a city budget?

In a technologically progressive organization, budgeting often lays out the blueprint for innovation over time. Mayor Mahan has set five clear focus areas: reducing street homelessness, increasing public safety, increasing beautification, increasing housing production, and attracting investments in jobs. Many of our boldest innovation investments track to those five areas, which have also been approved by the City Council, which creates a sense of shared focus and commitment.

Tactically, we often favor small-scale pilots before adding large technical debt to our organization. My team and I constantly meet with vendors to understand the latest technology and offerings, and we often seek to partner with other public sector agencies like the Valley Transit Agency to share costs and expand the scope.

What does "innovation" mean in a city government context, where residents expect stability?

"Innovation" can absolutely be a subjective term, but I would argue in San Jose, the bar is high. When you live in the same town as the headquarters of Zoom, Adobe, and one town over from Nvidia, your residents expect services to get better over time through technology. At any given time we have multiple pilots going on across 19 departments.

All of that being said, we do need to get the basics right. I applaud Mayor Mahan for having a commitment to performance management to ensure we are improving at what matters over time. Our City Council Focus Area Dashboards are a unified way of measuring progress on the five areas I previously mentioned. Ultimately innovation means safer communities, a reduction in the number of our unsheltered neighbors with pathways to employment, access to high-quality paying jobs and education regardless of the zipcode you were born into, clean and beautiful communities, and a supply of reasonably priced housing.

PII’s Notes: This definition reframes innovation away from technology for its own sake and toward tangible improvements in residents' lives, a crucial distinction that many civic tech initiatives miss. It also reveals the political reality: innovation must serve the administration's stated priorities or risk becoming an isolated experiment without institutional support.

Technology and Cities

Where do you see the most immediate opportunities for technology to improve urban life in San Jose?

Transportation - We have a city of just under 1,000,000 people and 182 square miles of land that ranges from foothills to a vibrant downtown area. Even modest improvements in transportation immediately impact thousands of residents.

Public Safety - Decreasing law enforcement and fire response times as we deal with staffing shortages is a primary goal. Additionally, disaster preparedness and mitigating fire risk is a significant concern given recent developments. Similarly, reducing traffic accidents as we strive for Vision Zero is another significant goal.

Economic and Workforce Development - Even in the heart of Silicon Valley we need to be vigilant in ensuring we are actively employing business development strategies to keep our local ecosystem full of diverse sectors and thriving. Further, we need to reduce friction for job seekers by building bridges between them and employers. We seek to provide a baseline level of AI literacy and resources so our residents remain competitive.

Outside of technology, we have 60,000 small businesses in San Jose with owners of different native languages, country of origin, and experience with running a business.

How do you evaluate whether a technology is truly solving a civic problem versus adding complexity?

Our first step is to often speak with the subject matter experts within the departments as a litmus test. Some solutions are objectively great offerings, but it may be the wrong time for San Jose or we may lack the human capital to effectively deploy it. Speaking with department heads and the City Manager's Office can provide vital strategic guidance as to the "when" and "how" of a deployment. Speaking with the front line workers who would be directly operating or benefiting from the system helps validate the problem statement the solution seeks to address, and helps us tailor the pilot for maximum efficiency.

The next step is a comparison of all other relevant investments and corresponding theories of change, to determine where the proposed efforts should fall in a list of prioritization.

PII’s Notes: This evaluation framework reflects a mature approach to technology adoption: it acknowledges that timing and organizational readiness matter as much as the technology itself. The emphasis on frontline worker input is particularly notable: many failed civic tech projects stumble because they optimize for executive dashboards rather than the people who actually use the tools daily.

What are the risks cities should be more vigilant about as AI and data-driven systems move deeper into public operations?

Increased technical debt - Certain tools like chatbots require a significant amount of backend configuration to be able to draw from relevant knowledge sources. Should the public agency seek to end the relationship, there must be a clear path to unwinding the integration.

Relevant training data to ensure there is similarity between the training data and the deployment environment. In San Jose, any solution must consider our Vietnamese, Mexican, and Japanese community members.

Vendors over-relying on non-proprietary applications - Should a vendor effectively be what is known as a "ChatGPT Wrapper," they likely have reduced control over their own infrastructure and service or quality of service may be negatively impacted by changes outside of their own control.

Cybersecurity - Governments hold a high amount of sensitive data and any endpoint that connects to internal systems must undergo thorough vetting. Failure to do so can result in data breaches of residents, lawsuits and penalties, and ultimately a loss of trust.

PII’s Notes: These concerns reveal the fundamental asymmetry between cities and startups: a venture-backed company can pivot away from a failed product, but a city that deploys a flawed system faces years of technical debt, potential lawsuits, and erosion of public trust. The emphasis on cultural competency in AI training data is especially relevant in San Jose, where 40% of residents speak a language other than English at home.

Working with Startups and Founders

What do startups often misunderstand about working with cities?

The Government as a Buyer often has completely different purchasing behavior compared to private industry. A few examples include:

• Sellers must be familiar with an agency's competitive purchasing thresholds, budgeting process, and standard contractual terms to be able to chart a path forward.

• Even free pilots may have to undergo legal, privacy, and cybersecurity reviews.

• Interest from a political leader like a Mayor does not equate to a guarantee that a City will buy your tool.

• The number of explicit GovTech companies and companies that sell to the government has significantly expanded over the last few years. Your idea is unlikely to be the first time that a government technology worker has heard of the premise—you must win on execution, specific functionality, or operating terms to stand out from your peers.

PII’s Notes: This reality check is essential for founders who assume that enthusiasm from a mayor or council member translates to a closed deal. The municipal procurement process involves multiple stakeholders including legal teams, IT departments, budget offices, and elected officials, each with veto power and different concerns. A founder who doesn't understand this will consistently misforecast sales cycles and burn through their cash runway waiting for deals that never materialize.

From your perspective, what makes a startup a strong partner for a city government?

• Demonstrated success in other jurisdictions of our size or makeup

• Strong customer service

• Flexibility in contractual terms

• Clear statements of efficacy and data

• Strong ethics, data privacy, and cybersecurity

• Commitment to accessibility and inclusivity of non-native English speakers

• If customer-facing: mobile-friendly environments, strong documentation and customer experience analysts

  
  

Cities are cautious adopters. How can founders align their timelines and expectations with public-sector decision-making?

Familiarize themselves with the budget process and key milestones. Generally speaking, begin advocating one year in advance of when you'd like to close a deal so it doesn't seem rushed. Reach out to folks who have been awarded contracts for guidance.

It's also key to have a clear understanding of the goals of elected officials and other stakeholders like the City Manager. For example, the difference between "solving homelessness" and "solving unsheltered homelessness" is pretty large, even though there's just an addition of one word.

When pitching to a city, what are the one or two things that instantly increase credibility?

• Recommendation from a trusted colleague who is either in or recently served in government as they understand the nuances of the demands we are placed under.

• Previous positive engagements with a major US city where there was a case study with post-implementation data included. Partnerships with universities to measure impacts also help credibility.

  
  

If you could give early-stage founders one piece of advice before they ever reach out to a city, what would it be?

Gain an understanding of the issue from the City's perspective (through published memos, speeches, and reports), the resident's perspective (through community meetings, public comment, and social media), and private industry (which players dominate and lead the industry, how is the same issue addressed by private actors).

PII’s Notes: This advice cuts to the heart of why so many civic tech ventures fail: they build solutions for problems they haven't fully understood. A founder who shows up having done this homework immediately distinguishes themselves from the dozens of others making cold pitches based on assumptions rather than ground truth.

Looking Ahead

Where do you see San Jose—and cities more broadly—five years from now in terms of innovation capacity?

I predict several developments to permit the advancement of more nimble procurement mechanisms for cities:

• Increased multi-agency procurements

• Streamlined procedures for pilots—specifically SaaS pilots

• Standardized forms like the GovAI Coalition's Vendor Fact Sheet and the recently released Data Sharing Agreement template

I predict greater investment in Privacy, AI, Innovation, and Cybersecurity roles in smaller and mid-size cities.

I envision a more tech-literate and empowered constituency which will permit more advanced lines of communication.

What's one area of city life you wish innovators would tackle more boldly?

Assist public agencies in understanding evidence-based differences in resource allocation based on geographic or community differences.

San Jose households have a wide variety of incomes and come from different areas of socioeconomic backgrounds. Individual neighborhoods have specific makeups, cultural assets, and levels of relative wealth. How can the city ensure equitable deployment of resources taking into account historic investment, geographic risks, or other key factors?

Help cities understand interventions they can make to prevent worker displacement and promote upskilling for the general public.

PII’s Notes: These challenges represent the next frontier of civic innovation: not just making government more efficient, but making it more equitable. The first challenge of understanding geographic disparities in need and investment, requires sophisticated data analysis combined with deep community engagement. The second acknowledges that as San Jose's economy evolves, the city has a role in ensuring residents aren't left behind by technological change, an especially pressing concern in a region that both drives and is disrupted by that change.

If you had an unrestricted "innovation budget" for San Jose, what's the first thing you'd invest in?

A discretionary budget for individual departmental pilots and innovation manager roles in each of the 19 City Departments.

Budget cycles are often on an annual basis and, after years of deficits, it's particularly difficult to preserve funds for a non-specified purpose when other clear needs exist. Preserving the ability to engage in pilots allows cities to capitalize on novel opportunities and reduces administrative burden related to execution. It also helps executives avoid an opportunity cost in electing to take on a new project.

Some departments have an existing designated manager like the Department of Transportation; others have people who informally fill the gap. If we had a key point person in every department, we could guarantee regular review of market offerings with greater inventorying of presentations, better data sharing between departments about pilots that would touch multiple departments, and the annual formation of a 2-year AI strategy.

PII’s Notes: This final answer reveals what's often missing from conversations about civic innovation: not breakthrough technology, but basic organizational capacity. Caines is describing the infrastructure layer that would allow San Jose to consistently evaluate, pilot, and scale new approaches across all city functions. It's an unglamorous answer that won't generate headlines, but it's precisely the kind of institutional investment that separates cities that occasionally experiment from those that systematically improve.

Key Takeaways for Founders and Civic Innovators

1. Align with political priorities: Innovation efforts without clear connection to leadership's stated goals risk becoming isolated experiments.

2. Plan for 12-18 month sales cycles: Cities move on budget timelines, not startup timelines. Factor this into your runway calculations.

3. Prove it elsewhere first: Case studies from comparable jurisdictions carry more weight than promises about what your technology could do.

4. Design for diversity: In cities like San Jose, solutions must work for multilingual, multicultural populations or they don't work at all.

5. Build for exit: Cities need clear paths to discontinue tools that don't work. Reducing technical debt and lock-in increases adoption likelihood.

6. Talk to frontline workers: The people who will actually use your tool know whether it solves a real problem. Executives can only guess.

The gap between Silicon Valley's innovation culture and the city government's operational reality is not just about speed, it's about fundamentally different constraints, stakeholders, and definitions of success. Caines' approach in San Jose offers a model for how cities can thoughtfully adopt new technologies without sacrificing stability, equity, or fiscal responsibility. For founders, it's a masterclass in what cities actually need versus what they're constantly pitched.