Business Intelligence Analyst & AI Automation Engineer

I turn messy data and manual work into systems people actually use.

Four+ years building data warehouses, BI systems, and production AI agents. I've worked with data over 10 million records at scale. Most of my work happens inside real client engagements, working with stakeholders who needed actual answers, not dashboards nobody opens.

Favour Osemujiamhe Atane
Favour Atane
Lagos, Nigeria Remote
10M+
records processed at scale in production
2M+
daily streaming events across pipelines
99.9%
pipeline uptime with automated validation
85%
manual processing time reduction via automation

How I work

Three things stay true whether I'm modeling a warehouse or building an agent.

01

I sit with the people using it

Requirements gathered secondhand produce dashboards nobody opens. I'd rather be in the stakeholder meeting than working off a spec.

02

I notice the problem before it's assigned

Most of what's in this portfolio started as something I flagged myself. A report eating two weeks, a join nobody wanted to touch. I spot the bottleneck and fix it.

03

I leave it usable without me

Reusable views instead of one-off queries. Self-service tools instead of standing requests. Documentation instead of tribal knowledge.

Two sides of the same work

Most projects below are client work. I've named the ones I can. The rest I describe in full detail with the client names removed.

Analytics & BI

Data warehouses, dashboards, and reports people trust

Medallion architectures, streaming pipelines, and BI delivery across Power BI, Snowflake, Sigma, and Domo — built for stakeholders who need an answer, not a spreadsheet.

View the work →
AI Automation

Agents that replace the manual work, not just describe it

Multi-agent pipelines, document extraction, and LLM-backed classification, built on n8n, Make.com, and custom orchestration — evaluated on cost, accuracy, and speed, not hype.

View the work →