All work Case study

Vennote Technologies

Backend for MTN's customer-facing digital assistant, Zigi, and the internal AI playground the team uses to curate and test it.

Role
Software Engineer, Backend
When
Jun 2025 – Present
Where
Lagos, Nigeria
Stack
Python, TypeScript, Pytest, OpenTelemetry, Grafana, CloudWatch
How the pieces fit: customer channels reach Zigi through MTN's MADAPI platform; the AI playground manages the knowledge base and response testing.
How the pieces fit: customer channels reach Zigi through MTN's MADAPI platform; the AI playground manages the knowledge base and response testing.

The context

Zigi is MTN's customer-facing digital assistant. My work sits behind it: the backend services that connect it to MTN's MADAPI platform, and the internal tools the team uses to keep its answers accurate. [ADD: team size and who you worked with]

What I built

  • Engineered scalable backend services for Zigi, integrated with MTN's MADAPI platform.
  • Built an internal AI playground for knowledge-base curation, content management, and text and voice response testing.
  • Developed and maintained Pytest unit and integration suites with high coverage.
  • Added structured logging and monitoring with OpenTelemetry, Grafana and CloudWatch.
  • Enforced coding standards through peer code reviews, quality checks and engineering documentation.

A hard problem

[ADD: one tricky problem you solved here: what broke or was unclear, the options you weighed, what you chose and why. Two short paragraphs is plenty.]

Results

  • About 40% more efficient knowledge-base and response-testing workflow after the AI playground launched.
  • Production issues are easier to diagnose thanks to structured logs, traces and dashboards.

What I'd do next

[ADD: one improvement you'd make with more time. It shows judgment.]

Next case study Mezovest Limited