Hi, I'm Santiago Díaz de Valdés Williamson.
CTO & AI Architect — Santiago, Chile
Electrical engineer turned AI builder. I've led a $2.2M government tech platform, shipped AI SaaS products at sub-$1/month infrastructure cost, and designed IoT hardware still running 9 years later. Based in Santiago — working globally.
~~~ Core expertise
Building AI from first principles — custom retrieval pipelines, multi-agent orchestration, cost-differentiated LLM architectures. Production RAG with hybrid retrieval, semantic chunking, and custom citation matching algorithms.
Cloud-native systems that scale without bleeding cash. 40+ AWS services in production, IaC-first, sub-$1/month fixed-cost designs, multi-account governance for enterprise complexity.
Bringing methodology where none exists — organizational diagnosis, SAFe introduction, team restructuring under delivery pressure. Led 15-20 engineers through a $2.2M project delivered on time.
End-to-end product delivery — from schema to UI to deployment. Modern React frontends, TypeScript-first backends, serverless-native data layers. Built and shipped solo in production.
Distributed system patterns with Domain-Driven Design and security-by-design. Multi-layer security from prototypes: rate limiting, guardrails, IAM least-privilege, prompt injection protection.
Large-scale data lakes and GIS systems integrating hundreds of sources. Designed a 40TB+ data lake with 262 geospatial layers from 50+ institutions. OGC protocol expertise at production scale.
~~~ Sectors
Full-stack engineering, architecture, and technical leadership end-to-end — from IoT hardware to AI SaaS, built as Tech Lead, Fractional CTO, or founder.
Production delivery in hierarchical environments — public procurement, ISO / GDPR compliance, and multi-institutional architecture.
Research-to-production with academic teams and PhDs — ANID / FONDEF programs and UC tech transfer.
~~~ What I do
Your startup needs technical direction but can't (or shouldn't) hire a full-time CTO.
You want to integrate AI into your product but don't know where to start, or your PoC doesn't scale to production.
Your tech project is behind schedule, your team isn't delivering, or your legacy system can't handle growth.
Solutions Architect
Developer
Cloud Practitioner
AI Practitioner
Well-Architected
~~~ Success stories
Featured
A ~$2M smart-city program for Santiago's Metropolitan Region.
Co-developed by Gobierno Regional Metropolitano de Santiago and UC Chile. Led 15–20 engineers across 7 interoperable sub-products spanning security, mobility, risk management, and territorial planning — delivered and handed over in 19 months.
Organizational dysfunction, unclear technical direction, a team working at cross-purposes. Government stakeholders with competing priorities. No shared methodology. No formal architecture. 19 months and $2.2M to deliver a working platform.
Chief Technology Officer. Not consulting — owning. Responsible for technical architecture, product strategy, team structure, hiring, vendor relationships, budget, and delivery.
Featured
Turned an academic project into production software serving 11+ municipalities.
Multi-agent AI that evaluates 400 CVs so hiring teams can focus on the top 5.
Recruiting teams waste enormous time on initial CV screening. Most AI tools are black boxes — candidates can't trust the criteria and managers can't audit the logic.
A production RAG system that runs at under $1/month. Full-stack AI academic assistant with verified citations.
Students struggle to navigate dense academic material. Existing AI tools hallucinate citations or fail on complex documents. The challenge: build something academically rigorous at a cost that supports a freemium model in a market of 11,000+ schools in Chile.
Sole designer, architect, and engineer. No team. I defined the product, designed the system, wrote the code, deployed the infrastructure, and conducted retrieval quality benchmarks.
Reduced measurement error from 30% to 0.25%. Device still running 9 years later.
Tu Interrogador — Bar Exam Prep Frontend (2019)
Frontend for a Chilean law exam prep platform. Custom recursive accordion optimized for mobile with hundreds of nested elements.
Vue.js 2 · JavaScript · Adobe XD · Sentry
Muvix — Cinema Ticketing Backend (2018)
Complete backend for a cinema chain. SOAP integration, payment gateway, QR tickets. Still operating.
Django · Celery · Redis · Docker · AWS
Sembrando Vitalidad — E-commerce (2022)
Custom e-commerce with complex product rules, SSR for SEO, full admin backoffice.
Next.js · Node.js · Express · MongoDB · AdminJS
Birthminder — Mobile Reminder App (2026)
Birthday reminder app with Facebook import, WhatsApp integration, cloud sync, bilingual.
React Native · Lambda · DynamoDB · EventBridge · CDK
And many more…
Prototypes, consulting engagements, and shipped-and-sunsetted products across nearly a decade of software work.
Ask me about them~~~ Process
Structured conversation about your technical situation — what's working, what isn't, what you're trying to achieve. You leave with clarity on where the leverage is.
Technical assessment: architecture review, team evaluation, or infrastructure audit. I deliver a written recommendation with prioritized actions.
I embed at the appropriate depth — strategic advisory, hands-on architecture, or interim technical leadership. Weekly touchpoints with async in between.
I build to hand over. Everything I create — documentation, architecture decisions, team structures — is designed to be owned by your team after I step back.
Monthly Retainer
Fractional CTO: 10-20 hrs/week
Project-Based
Architecture, AI system: defined scope
Advisory
Board or investor preparation: flexible
Not a fit if:
FAQ
A fractional CTO provides part-time or project-based technical leadership for startups and growing companies — without the cost of a full-time executive hire. You get CTO-level strategy, architecture decisions, and team leadership at a fraction of the commitment. It's ideal for early-stage startups that need senior technical direction but aren't ready for a full-time hire.
Both. I built AskMyCourse and CandidatePilot entirely solo — architecture, code, deployment, benchmarks. But I also led a 20-person multi-team organization at CEGIR under government accountability. I scale my involvement to what the engagement needs: hands-on when building, strategic when leading.
Yes. My AI systems operate at under $1/month in fixed infrastructure costs with 90%+ gross margins. I design for cost efficiency from day one — not as an afterthought. If you're worried about cloud bills eating your runway, that's exactly the kind of problem I solve.
I'm based in Santiago, Chile (UTC-4). I'm flexible to overlap with North American and European time zones — most of my clients are remote-first. I prefer weekly sync calls with async communication in between.
Yes. I was CTO of CEGIR, a $2.2M government smart city platform delivered to the Metropolitan Regional Government of Chile. I ran a formal RFI across 4 cloud providers, managed 8 concurrent vendors, and completed full governance documentation and operational handover.
Some repositories are public on GitHub. For private projects, I can walk you through architecture decisions, system diagrams, and technical depth without exposing source code. The case studies on this site are designed to show how I think and decide, not just what I built.
~~~ About
I'm a software architect, AI engineer, and technical leader based in Santiago, Chile. I hold a degree in Electrical Engineering with a specialization in Robotics and Automation from Pontificia Universidad Católica de Chile — the highest-ranked university in Latin America (QS #1). I graduated in the top 10% of my class and hold five AWS certifications.
I started programming at age 11. Since then I've worked across IoT hardware, cinema ticketing systems, EdTech, government platforms, and AI-native SaaS. I've been a backend developer, a hardware department head, a de facto architect, and a CTO — sometimes in the same year.
I build things to last. The IoT hardware device I designed at 26 — PCB, firmware, and cloud — is still in production 9 years later with no significant changes. The AI systems I build operate at sub-$1/month fixed infrastructure costs — not because I'm cheap, but because well-designed systems don't waste resources.
I work best at the intersection of ambiguity and high stakes. If you need someone who can walk into a technical organization, understand what's actually broken, and build a path forward — that's my zone.
~~~ Get in touch
I'm currently open to fractional CTO engagements, technical co-founder conversations, and select consulting projects in AI architecture and cloud infrastructure.
stgoddv@gmail.com LinkedInI respond to every serious inquiry within 24 hours.