Basil Ogbonna

Data Engineer

Career signalWeb DeveloperSoftware EngineerLead Software EngineerData EngineerSenior Data EngineerAI Engineer

About

Software taught me structure. Data taught me scale. AI taught me to keep asking better questions.

I am a Senior Data Engineer focused on making complex data systems reliable, reusable and easier to operate. I enjoy designing the foundations behind analytics: robust pipelines, clear models, trustworthy validation and deployment workflows that help teams move faster. I also build independent AI products to deepen my hands-on experience with agents, tools, human oversight and multimodal applications.

I care about the full path from a difficult problem to a system people can trust and use.

Engineering process

The work behind the outcome.

Small, inspectable views into how I approach AI, data, and software systems. These are process patterns—not performance claims.

Turn an uncertain question into an evaluated, observable system.

  1. Frame
  2. Retrieve
  3. Generate
  4. Evaluate
  5. Observe

ai.pipeline

context = retrieve(question, portfolio_facts)answer = model.generate(context, constraints)score = evaluate(answer, sources, risk)return answer if score.passes else abstain

Experience

The journey, company by company.

A brief of every chapter is always visible. Open one to see the work behind it.

Selected projects

A few things I've made curious on purpose.

Working systems and technical experiments, each with a public path to inspect or try it.
WhereToCode project cover

SoftwareContributor

WhereToCode

A web application that helps developers find suitable places to work from.

  • PostgreSQL
  • Express.js
  • React.js
  • Node.js
Lambdadoor project cover

SoftwareContributor

Lambdadoor

A Glassdoor-style product for Lambda students, built by a six-person team in eight weeks.

  • PostgreSQL
  • Express.js
  • React.js
  • Node.js
Matrades project cover

AIpersonal

Matrades

An active AI-assisted, multi-agent decision-support platform for market research, strategy, risk, and trading.

Portfolio, as proof

This portfolio is part of the work.

It is a working software, data, and AI product—not a static résumé dressed as a website.

  1. Structure

    Career history, projects, skills, and outcomes are modelled as connected data.

  2. Curate

    Only reviewed facts are projected into the portfolio experience.

  3. Match

    A role description is compared with published facts to explain fit.

  4. Answer

    The portfolio assistant answers from portfolio facts and returns supporting references.

Technology

Built as a real product.

  • Next.js
  • React
  • TypeScript
  • Supabase
  • PostgreSQL
  • pgvector
  • Inngest
Architecture & AI process
  • Private source material is ingested, chunked, and embedded before it becomes searchable portfolio context.
  • Hybrid retrieval combines lexical search with vector similarity, then ranks context before answer generation.
  • Public projection and Supabase row-level security ensure only facts selected for publication reach this portfolio.
  • Async workflow runs keep ingestion, embedding, review, and publication observable and recoverable.

Private workspace

More exists behind the public proof.

This page is the safe public surface. Owner-only tools remain private so personal context, preparation, and operational work stay protected.

  • Career knowledge base and fact curation
  • Job, application, and document workspaces
  • Interview preparation assistant and interview runs
  • Automations, analytics, and owner-only settings

Live modules

Test role fitAsk the portfolioSource link is not currently published.

Interactive proof

Ask Basil

portfolio facts ready

Ask about my engineering background, career journey, projects, outcomes, or role fit. I answer from the facts shared in this portfolio.

Tool trace waiting
  1. classify_questionawaiting question
  2. retrieve_portfolio_factsidle
  3. compose_grounded_answeridle

Start a conversation

Let's build something useful.

Have a data platform, engineering, or AI challenge worth solving?

Send me a note directly at

ogbonna.basil3@gmail.com