Independent data & policy practice

I build data systems and evidence for complex decisions.

Forecasting, applied AI and impact evaluation for organizations working with real operational and institutional constraints.

Based in Santiago, working across public policy, operations and applied data science.

01

Selected work

Systems built where data, institutions and day-to-day decisions meet.

02

Capabilities

Focused technical work, connected to the decision it needs to improve.

01

Applied AI for operational workflows

Documents → classification → human review

PYTHON · LLM APIS · RAG
02

Forecasting and decision systems

History → forecast range → planning decision

DATABRICKS · PYSPARK · SQL
03

Policy analysis and public systems

Data → evaluation → policy evidence

EVALUATION · SURVEYS · GOVTECH
04

Software delivery, orchestrated

Scope → partner build → owned data layer

AWS · R · SHINY · PYTHON

03

Built for real operating conditions.

A practical path from an ambiguous problem to a system your team can keep using.

  1. 01

    Define the decision

    What changes if the analysis is right, and who acts on it.

  2. 02

    Build the evidence

    Pipelines, models and indicators fitted to the sources that exist.

  3. 03

    Test under constraints

    Legacy systems, missing data and institutional timelines — before delivery.

  4. 04

    Deliver usable systems

    Documented and maintainable by your team, not by a slide deck.

Alfredo IñiguezSantiago, Chile · 33.45°S

04

Alfredo Iñiguez

Economist and data scientist. I spent four years inside the Chilean Ministry of Economy leading SUPER, the state digital permitting platform, and worked at FAO United Nations building models to evaluate agricultural policy.

That background shapes how I consult: I have shipped systems inside institutions, so I design for the constraints — legacy data, procurement and political timelines — that technical work often ignores.

2026— Astra AI · Data Scientist2025— OTEI, U. de Chile · Researcher2022–25 Min. Economía · Product Owner, SUPER2021 FAO UN · Technical Advisor
Curriculum (PDF)

05

Tell me about your problem.

A dataset that no one trusts, a forecast that keeps missing, a process you need evidence to change — start there.