Company Overview #
Core-AI is a Generative AI engineering practice in Montreal. We build secure, production-ready AI systems for law firms and engineering firms — typically 5 to 50 people — that want modern AI without handing their clients’ files to a third party.
That focus is deliberate. Both kinds of firm carry real confidentiality obligations, both are covered by Quebec’s Law 25, and both spend their days in documents — which is exactly where this technology earns its keep. A firm of that size doesn’t need a vendor running a playbook; it needs an engineer who understands the work.
Unlike service providers that rely exclusively on external APIs, Core-AI prioritizes in-house AI deployments using modern Large Language Models (LLMs). Your documents, your queries, and the model itself run on hardware you control. That preserves privacy, keeps data inside Quebec, reduces long-term operational costs, and avoids vendor lock-in.
Mission
Enable organizations to safely adopt Generative AI through reliable, private, and scalable AI systems that operate within their own infrastructure.
Vision
To become a trusted partner for businesses seeking long-term ownership of their AI capabilities, empowering them to build sustainable systems without external dependency.
Leadership #
Francisco Perdigon Romero
Founder & Principal AI Engineer
Francisco designs and deploys agentic AI, Retrieval-Augmented Generation (RAG), and LLM-powered systems for enterprise environments. He currently serves as AI Lead & Data Scientist at Ericsson, where his work focuses on agentic AI systems for 5G and LTE networks — experience he brings, in a personal capacity, to Core-AI's private AI deployments.
Before moving into industry, Francisco spent years as a deep learning researcher at École Polytechnique de Montréal, publishing work on medical image analysis and ECG signal processing. His research has accumulated 890+ citations and an h-index of 14.
He is also the creator of DeepFilter, an open-source ECG baseline-wander removal tool used by researchers and engineers in the signal-processing community. Francisco holds an MSc in Electrical Engineering (GPA 9.43/10) and graduated summa cum laude with his bachelor's degree in engineering.
Connect on LinkedIn →Strategic Objectives #
Accelerate Adoption
Move from AI experimentation to real operational systems that deliver measurable business value.
Promote Sovereignty
Enable companies to deploy AI while maintaining 100% ownership of their data, models, and infrastructure.
Production-Ready Solutions
Build AI systems designed for high reliability, scalability, and seamless enterprise integration.
Flexible Architectures
Design platforms that allow clients to combine local models with external services when necessary.
Core Principles #
Privacy First
Client data remains under the organization’s control at all times.
Practical AI
Focus on systems that deliver operational value, not just prototypes.
System Ownership
Clients gain the ability to operate and extend their systems independently.
Flexible Strategy
Support for both open-weight models and commercial providers.
Enterprise Integration
Seamlessly connect AI with existing enterprise tools and workflows.
Scalability
Architectures built to grow alongside your business requirements.
Technology & Engagement #
Our technical approach leverages Retrieval-Augmented Generation (RAG), AI Agent frameworks, and Vector Databases deployed via scalable microservices. This ensures your AI platform is as robust as any other piece of core IT infrastructure.
We work through a structured four-step process:
- Assessment — Identifying high-value AI opportunities.
- Design — Crafting a custom system architecture.
- Prototype — Rapid development for internal evaluation.
- Deployment — Scaling to production on your infrastructure.
Explore Our Detailed Services →
Value Proposition #
Core-AI helps organizations deploy powerful AI capabilities without sacrificing privacy, control, or long-term flexibility. We empower businesses to adopt Generative AI as a strategic internal capability rather than a temporary external tool.