Questions about freelance AI development and consultancy in Barcelona and Spain.
Yes. I am based in Barcelona and available for freelance AI development and consultancy engagements across Spain — including Barcelona, Madrid, and Valencia — and remotely throughout Europe. Typical engagements cover AI product development, automation, RAG systems, and technical project management.
Six core services: AI product development (scoping and building production-ready AI features), AI strategy and automation for SMEs and SMBs, multi-model AI routing (including AskARC integration), RAG-powered document intelligence systems, technical project management for AI projects, and full-cycle consultancy from initial discovery through to delivery.
I run a short discovery session to identify your highest-value automation opportunities, prototype the top use case, and take it to production. Focus is on practical, measurable outcomes — reducing manual overhead, accelerating workflows, or unlocking new capabilities — scoped to match your budget and risk appetite.
The most reliable path: (1) Audit your current processes for repetitive, rule-based tasks where AI adds clear value. (2) Start with one high-impact use case — typically document processing, customer communication, or internal knowledge retrieval. (3) Build a proof of value with real data before scaling. I guide clients through all three stages and handle the technical build.
Multi-model AI routing sends each query to the best AI model for that specific task rather than always using the same model. Simple queries go to smaller, cheaper models; complex reasoning tasks get frontier models. This typically reduces AI API costs by 40–80% with no quality loss. I built AskARC (askarc.app), an OpenAI-compatible gateway that does this automatically.
AskARC is an AI API gateway I built at FenxLabs. It is an OpenAI-compatible endpoint that routes queries across 200+ AI models based on complexity, cost, and privacy requirements. It delivers 40–80% AI cost reduction while maintaining EU-sovereign architecture. Operated by FenixMinds B.V. (Amsterdam) under GDPR. See askarc.app.
RAG (Retrieval-Augmented Generation) lets your team query a private knowledge base conversationally — like asking a search engine that reads your own documents and returns cited answers. Implementation: (1) Process and chunk your documents. (2) Embed them into a vector database. (3) At query time, retrieve the most relevant chunks and pass them to an LLM. Live demo at hiteshmankani.com/demos/dochat. Built using Azure OpenAI and MongoDB.
I scope MVPs tightly, prototype fast using real data, and prioritise production-readiness from the start — not throwaway demos. Every project includes architecture decisions, privacy-first design (especially for EU clients), and a plan for handover or ongoing maintenance. My background in technical project management means I also manage stakeholder alignment, timelines, and risk alongside the engineering.
I have delivered AI solutions across professional services (coaching, consulting, financial analysis), hospitality (WhatsApp booking and staff management systems), wholesale distribution (ERP systems), and field sales. The underlying patterns — document processing, automation, intelligent routing — apply broadly across industries.
A focused proof of value typically takes 2–4 weeks. A production-ready feature or small product takes 6–12 weeks depending on integrations and compliance requirements. Full AI product builds with multiple modules run 3–6 months.
For Barcelona-based small businesses, I offer: process audits to identify AI opportunities, affordable phased delivery (pilot first, then scale), Spanish-language support, local in-person workshops, and ongoing retainer arrangements. Engagements start with a free 30-minute discovery call.
Book a free 30-minute call at cal.eu/hiteshmankani/30min or email hitesh@fenxlabs.ai. I respond within one business day. Initial consultations are free.