LLM systems & agents
Agents, RAG and document pipelines for text-heavy work: legal drafting, meeting notes, OCR extraction and text-to-SQL over large warehouses.
Contact
A short note is enough: the problem, the data you have and your timeline. I reply within 24–48 hours, in English or Spanish.
The message didn't go through
Something went wrong while sending. Try again, or write to me directly at gairo@berkeley.edu. gairo@berkeley.edu
Services
Agents, RAG and document pipelines for text-heavy work: legal drafting, meeting notes, OCR extraction and text-to-SQL over large warehouses.
Site selection, market potential, POI matching and risk indices, delivered as interactive maps on MapLibre or Mapbox.
Demand forecasting, sales prediction and inference APIs, evaluated against a baseline before they reach production.
Pipelines, warehouses and dashboards on BigQuery, Redshift or PostgreSQL, with cost controls and privacy rules built in.
A defined scope, clear deliverables and a timeline. A good fit for taking a prototype to production.
Architecture and model reviews, stack and vendor choices and a second opinion before you commit budget.
Applied research on agents, knowledge graphs or spatial ML, with systematic evaluation, as in my medical AI work at UTSA.
Hands-on sessions for teams adopting LLMs or geospatial tools, drawn from training interns on production legal AI.