Daniel Reyes
Principal Engineer (AI/ML), YuSMP Group
Daniel Reyes leads applied AI and machine-learning topics for the YuSMP Group blog — retrieval-augmented generation, fine-tuning, AI agents, model selection, and the cost and ops of running LLMs in production. His articles are the decision frameworks our US and EU clients use before committing to an architecture.
- Born1985
- LocationBarcelona
- Experience15 years
- SpecializationApplied ML, LLMs, RAG & MLOps
- EducationPhD in Machine Learning, Universitat Politècnica de Catalunya
Principal Engineer · 12 articles
AI / LLM
12 min read
Machine Learning Software Development: How ML Systems Are Built in 2026
What machine learning software development is, how ML systems are built end to end, the process, tech stack, 2026 cost and timeline, MLOps and the main risks.
AI / LLM
12 min read
Generative AI Software Development: How to Build a GenAI Application in 2026
Generative AI software development in 2026 — architecture, model strategy (API, RAG, fine-tuning), the build process, tech stack, realistic cost and timelines.
AI / LLM
12 min read
AI Software Development: How Teams Actually Build With AI in 2026
AI software development in 2026 — where AI helps across the SDLC, the real adoption and productivity data, the risks, the tool categories, and how to adopt AI without shipping worse software.
AI / LLM
13 min read
How Much Does It Cost to Build an AI Chatbot in 2026?
What an AI chatbot costs to build in 2026 by type — rule-based, NLU and LLM/RAG — plus integrations, running costs and build-versus-buy.
AI / LLM
12 min
Generative Engine Optimization: how to get cited by AI search in 2026
How to get cited by Google AI Overviews, ChatGPT and Perplexity in 2026 — front-loaded answers, entity density, FAQ schema, the llms.txt reality and AI-visibility measurement.
Custom
12 min
Loan Origination Software Development
How to build loan origination software (LOS) — application intake, KYC, credit decisioning and risk scoring, underwriting workflow, loan management, build vs platform, cost and timeline.
AI / LLM
16 min read
Claude vs GPT-4o for product teams in 2026
Engineering comparison for production teams — SWE-bench / GPQA / agentic benchmarks, cost per 1M tokens, prompt caching, EU residency, when to choose which.
AI / LLM
20 min read
AI agents for enterprise in 2026
Production stack 2026: LangGraph, MCP, Anthropic SDK, Claude 4.6 / GPT-4o, orchestration patterns, cost-per-task economics, human-in-the-loop.
Enterprise
14 min
AI Integration in Enterprise Software 2026
How enterprises integrate AI into existing software in 2026 — use cases, embedding AI into ERP/CRM and legacy systems, governance, the EU AI Act, cost and ROI.
AI / LLM
18 min read
LLM fine-tuning cost benchmark 2026
H100/H200 GPU hours, dataset prep, LoRA vs full FT, RLHF, when to fine-tune vs use RAG — with three worked enterprise examples.
AI / LLM
17 min read
RAG vs fine-tuning in 2026 — what to choose
Decision guide for RAG, fine-tuning, and hybrid in 2026: Claude 4.6 / GPT-4o / Llama 4 / Mistral Large 3 benchmarks, cost per 1M tokens, retrieval stack, eval pipelines.
Custom
10 min
The Custom Software Development Process
The custom software development process explained — discovery, design, build, QA, launch and support: what each SDLC stage delivers, costs and how long it takes.
Building something in retrieval-augmented generation? Let's talk.
Response within 1 business day. NDA on request.
Get a proposal
Share a few details and a senior consultant will reply within one business day.