Jonathan

About

I build and study intelligent systems, and write about AI agents, memory, evaluation, open models and the engineering required to make AI work in the real world.

Jonathan Atiene

I'm an AI engineer, researcher and software builder interested in one broad question:

What does it take to turn increasingly capable AI models into systems people can actually depend on?

My path into AI has been unusually interdisciplinary. I originally studied human anatomy and neuroscience before moving into software engineering, where I spent several years building backend, frontend, cloud and production systems. That engineering foundation eventually led me deeper into machine learning, and I later completed an MSc in Artificial Intelligence. Today those fields overlap in my work. I'm currently based in London, building production AI systems at Hive Science.

I'm interested not only in what models can do, but also in the infrastructure around them — how we evaluate their behaviour, how they interact with existing software systems, and how humans interpret and make decisions with their outputs.

What I work on

Production AI systems

Building and studying the engineering layers required to move AI from prototypes into reliable software: agentic workflows, evaluation, context engineering, retrieval, memory, observability, structured generation, reliability, and human-in-the-loop systems.

Applied AI research

Practical machine-learning problems where model efficiency, intelligent systems and real-world deployment intersect. Neuroscience and behavioural science show up here as supporting lenses — useful when they genuinely inform how a system should be built, evaluated or trusted, not as ends in themselves.

How I think

I approach AI from three perspectives simultaneously:

  • Engineer — can this system be built, tested, operated and trusted in production?
  • Researcher — what evidence supports the assumptions we're making?
  • Product thinker — how will a human actually understand and use what the system produces?

The interesting AI problems increasingly need all three.

Selected experience

  • Senior Full Stack AI Engineer · Hive Science

    Building the AI platform end-to-end — LLM agent harnesses, context and memory architectures, and serverless data pipelines on AWS (Python, TypeScript, Pydantic, Qdrant, Terraform). Lifted processing throughput from ~20 jobs/month to 60+/week through automation and orchestration.

    Nov 2025 – Present

  • Lead AI Engineer · TrendMind

    Architected an AI platform end-to-end — structured content strategy, agentic workflows with retrieval, real-time engagement analytics, and human-in-the-loop approval. Lifted engagement metrics ~20%.

    Oct 2024 – Oct 2025

  • Senior Full Stack Engineer · CodeRabbit

    Optimized prompt templates and context packing to cut LLM cost ~15%. Migrated review configuration to YAML and built Jira and Linear integrations.

    Aug 2023 – Nov 2023

  • Senior Full Stack Engineer · Oaks Lab

    Led large features across five startup projects; re-architected the admin portal's front-end and APIs. Improved throughput ~30% and cut error rates ~20%.

    Jul 2022 – Feb 2024

  • Full Stack Engineer · Second Company

    React Native performance and native modules for iGarage; refactored auth and notifications; built a multi-tenant dashboard system.

    Sep 2021 – Jun 2023

  • Full Stack Developer (Lab Head) · Deposits

    Built Renapay's web app in Vue.js; rebuilt the invoicing system; automated white-label deployments through a custom CI/CD pipeline.

    Apr 2021 – Oct 2021

  • Full Stack Developer · Engage

    Built a drag-and-drop email template editor, shipped 2FA and multi-role access, improved SDK utilities and observability.

    Mar 2021 – Apr 2022

Education

  • MSc Artificial Intelligence · De Montfort University

    Deep Learning, Machine Learning, Data Mining, Evolutionary Computing, Fuzzy Systems, NLP. Thesis on efficient brain-tumour segmentation using parameter sharing and pruning in RAAGR²-Net.

    Feb 2024 – Jun 2025

  • BSc Human Anatomy / Neuroscience · University of Port Harcourt

    Anatomy, biochemistry, neuroscience, research methods, histology. Project on neural pathway mapping and cognitive function analysis.

    Jan 2014 – Dec 2019

A full CV is available as a PDF. For anything else, press & contact has bios, headshots and email.