I integrate AI into your SaaS without breaking your team's velocity
2-week diagnosis. Agents and process automation. AI features in product.
Backend dev with 8+ years in Go now helping LATAM founders integrate AI where it actually moves the needle — no chaos, no overhead, no empty 10x promises.
I've worked with startups and enterprises: Bayonet, Adevinta, Acklen Avenue, Manzana Verde, Banco Santander, Accenture.
You know you need AI — you don't know where to start
Here's what I keep seeing in SaaS founders across LATAM:
You want to use AI but don't know where it's a priority
Every conference talks about 'AI for SaaS'. None tells you which process in YOUR company will pay back the investment first. You end up paralyzed or building a chatbot that nobody uses.
Agencies charge for slides, not implementation
Quotes for 'AI strategy' come with 80-slide PowerPoints, 6-month timelines, and a price tag in the tens of thousands — before a single line of code. You need someone who ships.
Your team has no time to learn ML while shipping features
Your devs are heads-down on the roadmap. Asking them to learn LangChain, RAG, vector databases, and token economics on top of that is unrealistic. You need an outside specialist.
Three ways to start, depending on where you are
From a 2-week diagnosis to AI features running inside your product.
🔍 AI Diagnosis (2 weeks)
Audit of your processes and product to find where AI generates the highest ROI. Interviews with founders, leads, and ops. Deliverables: priority map (impact × effort), implementation roadmap, and cost estimate per use case. You leave with clarity and an action plan — even if you don't hire me for the next step.
⚙️ Agents & Process Automation
Automate workflows with LLMs and agents using n8n, LangGraph, or Claude/OpenAI APIs depending on what fits your team and budget. Integrated with your existing stack: Slack, HubSpot, WhatsApp Business, Postgres, Airtable. Built so your team can maintain it.
🚀 AI Features Inside Your SaaS
I build AI features inside your product itself: RAG over your own data, semantic search, copilots, content generation. Includes architecture decisions, token economics, monitoring, and graceful degradation. Code lives in your repo; I work with your devs.
Packages and investment ranges
USD. Fixed pricing by scope — not by the hour.
AI Diagnosis
Ideal for: founders who suspect AI can help but don't know where to start
- 1-2 weeks of work, async + 3-4 sync sessions
- Interviews with founders, leads, and ops
- AI opportunity map for your business (impact × effort)
- Implementation roadmap with phasing
- Per-use-case cost estimate (USD)
- Recommended stack (build vs buy, providers)
- 30 days of async follow-up support
Implementation
Ideal for: founders with a clear use case who need it built
- 4-8 weeks, depending on scope
- 1 agent or end-to-end automation in production
- Built with n8n, LangGraph, or custom APIs (we'll decide together)
- Integration with your existing stack
- Token & cost monitoring
- Documentation and training for your team
- 60 days of post-launch support
AI Features in SaaS
Ideal for: SaaS that want to ship AI features inside their product
- 8-16 weeks, depending on complexity
- Architecture and AI feature design
- RAG, copilots, semantic search, or generation as fits
- Integration with your real codebase (your repo, your stack)
- Token economics: realistic margin per user
- Observability: metrics, logs, alerts
- Pairing with your devs so they own the feature post-handoff
Frequently asked questions
Everything I get asked in the first call.
How is this different from hiring an AI agency?
An agency sends you a senior to the kickoff and juniors to the implementation. With me, the only person on the project is me — a backend dev with 8+ years in production. Fewer meetings, fewer middlemen, and an ICP-aligned price.
Do you work async or sync?
Mostly async with 2-3 sync sessions per week (calls/Slack-huddle). I work from Spain in CET, so I overlap nicely with LATAM working hours during their mornings. All deliverables are documented in writing — no value lives only inside someone's head.
What if my team has no AI experience?
That's the common case. Part of the engagement is leaving your team in a position to maintain what I build. Internal documentation, paired coding sessions, and code reviews. The goal is not to lock you into me forever.
Do you charge by the hour or fixed price?
Fixed price by scope. I tell you up front what's in the package, what isn't, and what costs extra. If we have to change scope mid-project, we sign an amendment in writing. No surprise invoices.
What if I'm not sure I need AI?
Then the 30-min call is free for exactly that reason. If after the call I think AI isn't the right answer for you right now, I'll tell you. I prefer not to take projects that won't deliver value — bad references travel further than good ones.
Can I see your previous work?
I publish open technical content on my blog (carlos.lat/blog) and share specific implementations on a call. Client work is under NDA, but I share generalized case studies once the work is in production.
Ready to move AI from your backlog to production?
30-minute call, free. I'll tell you if I can help — and if not, who to recommend.
Let's talk on WhatsAppOr email me at c.garcia@carlos.lat