PAULO VILA AI VERSION
McKinsey → BCG → Carrefour → Mastercard → Imaginos
I Build Systems. AI infrastructure architect and product executive with a 29-year track record spanning strategy consulting (McKinsey, BCG), corporate leadership (Mastercard, Carrefour), and hands-on systems engineering. Currently founding a local-first, multi-agent AI platform in Rust powering five live applications.
Combines board-level commercial judgment — $100M+ platforms, $30B+ merger delivery, 1M+ user launches — with the ability to design and ship the runtime personally. Bilingual English/Spanish, fluent French.
$100M+ B2B platform · 1M+ neobank users · $30B+ merger delivery · $20M mine turnaround · 60K-member executive community
- Shipped 5 production AI applications on a single Rust architecture — compliance/contracts (Vetra), algorithmic trading (Latinos), healthcare membership (Movilo), ecotourism/biodiversity (Garcero), executive portfolio (paulovila.org) — all sharing one orchestration, memory, routing, and governance stack.
- Reduced AI inference cost to $0 per query by deploying an on-premises GPU cluster (2× RTX 3090) for LLM, image generation (FLUX/ComfyUI), and multilingual embeddings — cloud APIs demoted to failover-only.
- Deployed 13 autonomous AI agents in 9 months with persistent scoped memory, semantic recall (multilingual embeddings), and multi-channel delivery (web, Telegram, WhatsApp) sharing one conversation context layer.
- Built a zero-shot NLP pipeline (GLiNER + GLiREL + GLiClass on GPU) for knowledge graph construction — entity recognition, relation extraction, and sentiment classification without labeled training data or per-domain fine-tuning.
- Implemented Model Context Protocol (MCP) bridges for Hera and Memento, and a self-correcting agentic coding loop: Hera edits Rust source, runs cargo_check, and iterates until the build is clean — autonomously.
- Engineered sub-millisecond IPC (1.019 ms avg, 1.136 ms p95 — 300×+ over prior Next.js stack) with automatic Postgres HA failover (~2 min RTO, RPO ≈ 0) via streaming replication and quorum witness.
Single-operator at platform scale: design the core abstractions, build them in Rust, ship to real users across web + Telegram + WhatsApp. Five apps (paulovila.org, vetra.trade, latinos.paulovila.org, movilo.club, elgarcero.com) share one runtime today.
- $0/query AI inference cost — on-prem GPU (2× RTX 3090) vs. ~$0.01–0.05/request on cloud APIs
- Rust warm-path latency: 1.019 ms avg, 1.136 ms p95 (Vetra-rust); 1.1 ms TTFB (Movilo) — 300×+ over prior Next.js stack
- 13 autonomous agents deployed in 9 months with persistent memory and semantic recall
- Zero-shot NLP pipeline: GLiNER + GLiREL + GLiClass (GPU) for knowledge graph extraction without labeled data
- Automatic Postgres HA: ~2 min RTO, RPO ≈ 0 via streaming replication + quorum witness
- 5 production verticals (compliance, trading, healthcare, ecotourism, portfolio) on one shared runtime
- $100M+ B2B credit card sourcing platform at Mastercard (18 months)
- Supported launch of Colombian neobank reaching 1M+ users
- 5× revenue growth for Mastercard Labs-as-a-Service in LAC
- 60+ innovation projects delivered, 90% client satisfaction
- 3 Mastercard President's Awards
Stanford University
The Wharton School
Mastercard, 2022 — women-led fintech LAC
Mastercard, 2020 — Digital Datathon LAC (best in company)
Mastercard — selected from ~30K global employees
Mastercard, 2016
Mastercard, 2016