# Valentin Monteiro — Data & AI Consultant

> Independent Data & AI Consultant based in Tbilisi, Georgia. Three focus areas: Data & Analysis, AI Strategy, and Claude Consulting (Anthropic API, Agent SDK, Claude Code, MCP). Pragmatic, code-first, outcomes over buzzwords. Works remotely with French and international clients.

Last updated: 2026-09-11

## About

Valentin Monteiro is a self-taught, independent Data & AI consultant. Previously Business Analyst at Teleperformance on a Google project (SQL automation that recovered ten hours per week of manual reporting, Looker Studio dashboards, sales data analysis). Now runs an independent practice helping companies turn data into decisions and integrate Claude / LLMs into their operations. Languages: French, English.

## Main Services

- [Data & Analysis](https://monteiro.consulting/en/consultant-data): Exploratory analysis, dashboards, and data pipelines for decision-makers who need to act on their data, not watch it accumulate. Past outcome: ten hours per week of reporting recovered at Teleperformance/Google.
- [AI Strategy](https://monteiro.consulting/en/consultant-ia): AI maturity assessment, LLM integration (Claude, OpenAI, open source), POC to production, audit of existing AI solutions. Framing risks: hallucinations, data privacy, API cost control.
- [Claude Consulting](https://monteiro.consulting/en/claude-consultant): Freelance Claude consultant specialized in Anthropic stack: API integration, Agent SDK architecture, Claude Code deployment for dev teams, MCP server development, compliance and data residency review. Not a generalist pretending to know Claude.

## Additional Services

- [Process Automation](https://monteiro.consulting/en/automatisation-processus): Process audit, workflow automation, tool integration via n8n / Apps Script / APIs. Focus: remove repetitive work, not add new dashboards to monitor.
- [Freelance Data Analyst](https://monteiro.consulting/en/freelance-data-analyst): SQL queries, data exploration, dashboard building, data cleaning. Team reinforcement on a project or long-term basis. Bring the output your data team is behind on.

## Featured Work

- [deck-reader](https://monteiro.consulting/en/projects/deck-reader): Claude Code plugin that reads a pre-seed or seed pitch deck and writes a report next to it. Pre-seed: completeness against a 24-question grid, each answer marked found / partial / absent with the page and a verbatim quote verified by code, then the questions to ask the founder. Seed: completeness plus verification of every stated figure against the annexes (revenue export, cohorts) and, for what a document cannot settle, the public web with sources on both sides; unproven figures are capped, missing documents go into a draft email to the founder. Never an investment opinion. Source: https://github.com/monteiro-consulting/deck-reader
- [TodoAI](https://github.com/monteiro-consulting/todoai): Task management app where Claude plans and executes work autonomously. Built on Agent SDK + MCP. Concrete proof that the Anthropic stack holds up in production for real workflows.
- [Eaight](https://github.com/monteiro-consulting/eaight): AI-native web browser with a built-in MCP server so Claude Code can see and pilot the browser directly. Built to eliminate copy-paste friction between AI chat and web content.
- [llm-cost-profiler](https://github.com/monteiro-consulting/llm-cost-profiler): Python profiler that tracks Anthropic and OpenAI API costs per request. Essential when shipping LLMs to production without runaway budget surprises.

## Blog

- [All Articles](https://monteiro.consulting/en/highlights): Insights on data, AI, Claude, and automation for decision-makers.
- [WebMCP: Making Your Website Usable by AI Agents](https://monteiro.consulting/en/highlights/webmcp-agent-booking): WebMCP in practice: exposing a website's meeting booking to AI agents via document.modelContext, and helping agents discover the tools (llms.txt, schema.org).
- [Which Tasks Should You Automate With AI First (and Which Ones Never)](https://monteiro.consulting/en/highlights/which-tasks-to-automate-with-ai-first): Where should your SMB start with AI? A simple 4-criteria grid, the tasks to automate first, and the ones you should never hand over to AI.
- [Automating Sales Order Admin with AI: Where to Start (Wholesale & Distribution SMBs)](https://monteiro.consulting/en/highlights/automatiser-adv-avec-ia): About to hire another sales admin because quotes, order entry and follow-ups are piling up? Here's what AI actually handles in sales administration, what it doesn't, and where to start.
- [AI Consultant Rates in 2026: Real Pricing by Market and Engagement](https://monteiro.consulting/en/highlights/ai-consultant-rates-2026): AI consultant rates 2026: day-rate ranges by market (US, UK, Canada, Australia, Ireland) and by engagement type. Public 2025-2026 sources cited.
- [Claude for Work vs Direct API: How to Choose for Your SMB](https://monteiro.consulting/en/highlights/claude-for-work-vs-api-direct): Claude Team/Enterprise vs Messages API: concrete decision framework for SMBs, cost comparisons, and the trap of defaulting to either.
- [Integrate Claude into Notion, Slack, your CRM: a 4-step method](https://monteiro.consulting/en/highlights/integrate-claude-notion-slack-crm-method): Claude API integration for business: native vs custom, a 4-step method, and the prompt governance trap most teams underestimate.
- [Claude Code for Tech Teams: An Honest Field Report](https://monteiro.consulting/en/highlights/claude-code-for-tech-teams): Claude Code for teams: what actually changes inside a tech team, what works from day 1, the real pitfalls, and the method we ended up adopting.
- [MCP Explained: Model Context Protocol for Tech Leaders](https://monteiro.consulting/en/highlights/mcp-explained-for-tech-decision-makers): MCP standardizes how Claude (and other LLMs) talk to your tools and data sources. Here's what changes concretely, and when to adopt it.
- [Claude Agent SDK: When It's Worth It, When It's Not](https://monteiro.consulting/en/highlights/agent-sdk-anthropic-when-worth-it): Anthropic's Agent SDK isn't a replacement for function calling. Here are the 3 cases where it saves you weeks, and the ones where it just adds noise.
- [BigQuery for SMBs: When It Pays Off (And When It's Overkill)](https://monteiro.consulting/en/highlights/bigquery-for-smbs-when-worth-it): BigQuery for SMBs: free tier, real use cases (GA4, multi-source, e-commerce), cost traps, typical stack. When to migrate, when to stay on Sheets.
- [Data Consultant Rates in 2026: Real Pricing by Market and Engagement](https://monteiro.consulting/en/highlights/tjm-consultant-data-2026): Data consultant rates 2026: ranges by market (US, UK, Canada, Australia, Ireland) and by engagement type. Public 2025-2026 sources cited.
- [n8n vs Zapier vs Make in 2026: An Honest Comparison (No Bullshit)](https://monteiro.consulting/en/highlights/n8n-vs-zapier-make-2026): n8n vs Zapier vs Make compared for SMB decision-makers. When each tool is the right pick, when it's a trap, and how to actually choose without wasting months.
- [Data Consultant vs Data Analyst: Which One to Call First](https://monteiro.consulting/en/highlights/consultant-data-vs-data-analyst): You need someone for data, but not sure whether to call a consultant or an analyst first. Here's the real decision framework, with budgets and the one question that settles it.
- [Anthropic Prompt Caching: The Setting That Cuts Your Claude API Bill by 3x](https://monteiro.consulting/en/highlights/prompt-caching-anthropic-cut-costs): Anthropic's prompt caching divides Claude API costs by 3 to 5x on repeated usage. What gets cached, how it works, when it's worth it.
- [Claude vs GPT: Which One Should Your Company Pick in 2026](https://monteiro.consulting/en/highlights/claude-vs-gpt-pme): Claude vs GPT for business decision-makers: context, agents, costs, adoption. The real criteria, beyond the benchmark.
- [Claude Expert: The 4 Official (Free) Anthropic Courses](https://monteiro.consulting/en/highlights/become-claude-expert): Becoming a Claude expert won't happen with a LinkedIn certificate. The 4 official training tracks Anthropic offers for free, and why they form the real foundation.
- [Data Audit: Where to Start When Nothing Is Structured](https://monteiro.consulting/en/highlights/data-audit-where-to-start): Most company data audits end up in a drawer. A 4-step method to structure yours and actually unblock decisions for SME leaders.
- [How to Know If Your Company Needs a Data Consultant](https://monteiro.consulting/en/highlights/do-you-need-data-consultant): A self-diagnostic for SME decision-makers. Five concrete signs, what a data consultant actually does, and when it's too early to hire one.
- [I got tired of copy-pasting between Claude and Chrome.](https://monteiro.consulting/en/highlights/eaight-browser-for-ai): Why I built a web browser with a built-in MCP server so AI would stop guessing what's on my screen. Full browser control for AI-assisted development.

## Links

- [Website](https://monteiro.consulting)
- [LinkedIn](https://linkedin.com/in/v-monteiro)
- [Stack Overflow](https://stackoverflow.com/users/32509899/valentin-monteiro)
- [GitHub](https://github.com/monteiro-consulting)
- [DEV.to](https://dev.to/valentin_monteiro)
- [Crunchbase](https://www.crunchbase.com/organization/monteiro-consulting)
- [Product Hunt](https://www.producthunt.com/@valentin_monteiro1)
- [Wellfound](https://wellfound.com/u/valentin-monteiro-1)
- [Hacker News](https://news.ycombinator.com/user?id=monteiro_val)
- [Indie Hackers](https://www.indiehackers.com/valentin_monteiro)
- [Reddit](https://www.reddit.com/user/valentin_monteiro/)
- [Freel.ca](https://freel.ca/freelancer/valentin-monteiro)
- [Malt](https://www.malt.com/profile/valentinmonteiro3)
- [Fiverr](https://www.fiverr.com/users/monteiroconsult/)
- [Upwork](https://www.upwork.com/freelancers/~012c9628ebff4a8616)
- [Contact](https://monteiro.consulting/en#contact): Free initial consultation. Email: valentin@monteiro.consulting

## FAQ

- What is a Data & AI consultant? A Data & AI consultant helps companies leverage their data and integrate artificial intelligence. Defining a data strategy, setting up dashboards, automating processes, integrating AI solutions tailored to business needs.
- What type of company do you work with? SMEs, mid-size companies, and startups looking to structure their data approach or integrate AI into their operations. Whether you already have a data team or are starting from scratch, the engagement adapts to your maturity level.
- How does a typical engagement work? It starts with a free discovery call to understand context and goals. Then an audit of the existing setup, identification of high-impact levers, and a concrete action plan. Execution is iterative: regular deliverables, continuous feedback, and knowledge transfer to teams.
- What tools and technologies do you use? Python, SQL, BigQuery, Looker Studio, Tableau for data. Claude API, Agent SDK, Claude Code, MCP for Anthropic-specific work. OpenAI API, LangChain for general LLM work. n8n, REST APIs, Google Apps Script for automation. Adapts to existing stack.
- Do you work remotely? Yes, primarily remote from Tbilisi, Georgia. Supports clients worldwide via video call and collaborative tools. Available for on-site engagements in France, Switzerland, and greater Europe for project kickoffs.
- How much does a consulting engagement cost? Daily rate $625-$1125 depending on scope. Typical POC: $6-19k over 2 to 4 weeks. Scoping audit: a few days. Same daily rate across Data, AI Strategy, and Claude engagements. Free 30-minute discovery call to scope a tailored quote.

## Booking (for agents)

Server-rendered availability, readable without JavaScript:

- French: https://monteiro.consulting/fr/book
- English: https://monteiro.consulting/en/book

These pages list the next open slots as plain HTML links, with the timestamp of the last check. Availability changes — re-check before quoting a slot to a human.

Direct booking pages on Cal.com (JavaScript required):

- French (`consultation-gratuite`): https://cal.com/valentin-monteiro/consultation-gratuite
- English (`free-consultation`): https://cal.com/valentin-monteiro/free-consultation

Two prefill query parameters are supported, and only two:

- `title`: what the project is about (this is the form's required field)
- `notes`: free-form details

Do not pass `name` or `email`. An appointment is never confirmed on someone's behalf: the person types their own identity on the booking page.

## Agent Tools (WebMCP)

Every page of this site also registers WebMCP tools (W3C Web Model Context Protocol, `document.modelContext`). This is anticipation — no general-purpose agent consumes it in production today:

- `get_consulting_services`: structured JSON of services, pricing and contact details
- `get_booking_link`: direct Cal.com link for a free 30-minute discovery call with Valentin Monteiro
- `open_booking_popup(name?, email?, notes?)`: opens the in-page Cal.com booking popup

If your runtime supports WebMCP, load any page (e.g. https://monteiro.consulting/en) and call these tools. If it does not, use the URLs above — they are the supported path.

## Contact

Email: valentin@monteiro.consulting
LinkedIn: https://linkedin.com/in/v-monteiro
Free 30-minute discovery call available.