Career Highlights

Current Role: Member of Technical Staff @ Cohere

Building and improving Compass, an enterprise-grade retrieval platform that powers RAG pipelines. Key contributions include:

Designed the end-to-end evaluation framework with retry mechanisms, checkpointing, and LLM-based evaluation for scalable IR experiments
Shipped multiple integrations: web2compass for scraping website into Compass, MCP tooling for North, and code2compass for code bases ingestion into Compass for semantic code search
Improved Compass Parser efficiency for vision-language models, achieving 50% memory reduction in image-to-markdown tasks
Led development of Compass SDK V2 with async support and built the Compass Asset Service for scalable ingestion pipelines
Implemented Postgres-based job queue replacing Celery for improved reliability in long-running jobs
Contributed to open-source by adding Cohere support to Pydantic-AI

Key Amazon/AWS Achievements

Founding engineer on Amazon Managed Apache Airflow (MWAA), leading design and launch at re:Invent 2020; scaled to 15+ global regions
Led modernization of the Distributed Job Scheduler backend using Elasticsearch, powering large-scale scheduling across Amazon
Built platform services for Amazon Mobile Shopping App serving millions of retail customers
Championed open source with amazon-mwaa-docker-images repository serving thousands of MWAA environments

Systems & Low-Level Programming Background

C++ expertise: Lead developer on Alusus Programming Language compiler using LLVM, imaging modules with AutoCAD DWG/DXF reverse engineering
High-performance systems: Financial workflow orchestration platforms, real-time alerting systems with Elasticsearch
Infrastructure at scale: Google App Engine migrations, NetworkedBlogs optimization for hundreds of thousands of users

Selected Projects

Compass @ Cohere - Enterprise search and retrieval platform
microtorch - Lightweight deep learning framework
neuroscout - LLM-powered NeurIPS paper analysis tool
image-search - Natural language image search using CLIP
pat-cli - ML-powered log clustering tool
amazon-mwaa-docker-images - Docker infrastructure for Apache Airflow
Alusus Programming Language - LLVM-based programming language compiler
Alkitab - Modern Quranic text processing toolkit

Skills & Technologies

Machine Learning: PyTorch, CLIP, BERT, LLM evaluation, LLM-as-a-judge, synthetic data generation, information retrieval, multi-modal models, autograd systems

Cloud & Infrastructure: AWS (ECS, Lambda, Step Functions, DynamoDB, RDS, CloudWatch), Elasticsearch, Docker, PostgreSQL

Programming Languages: Python, C++, JavaScript/TypeScript

Publications

Academic Publications

Alexa Visual Item Selection (AVIS) Dataset (Amazon Computer Vision Conference 2023)
Visual Item Selection with Voice Assistants (ACM Web Conference 2023)
Advanced Composition in Virtual Camera Control (International Symposium on Smart Graphics 2011)
Advanced Composition in Virtual Camera Control (MPhil Thesis, Newcastle University, 2011)

Technical Articles

Installing Jupyter on a Cloud Machine (Dec 2019)
Useful Shell Tools/Tips for Increasing Productivity (Dec 2019)
Full Text Search – Part 3: Building Inverted Index (June 2016)
Full Text Search – Part 2: Token Processing (May 2016)
Full Text Search – Part 1: Tokenization (May 2016)
Easy Caching with C# and PostSharp (Dec 2013)
Creating Strongly Typed Custom Collections in C# (Sep 2002)
Why is Psychology not Taught in Schools? (Nov 2019)
Focus: Reduction of Distractions (Jan 2018)

Reading List

2025

Good Energy: The Surprising Connection Between Metabolism and Limitless Health by Casey Means and Calley Means
The Whole-Brain Child: Revolutionary Strategies to Nurture Your Child’s Developing Mind by Daniel J. Siegel and Tina Payne Bryson
The Expectant Father: The Ultimate Guide for Dads-to-Be by Armin A. Brott
The Anxious Generation: How the Great Rewiring of Childhood Caused an Epidemic of Mental Illness by Jonathan Haidt
The Coddling of the American Mind: How Good Intentions and Bad Ideas Are Setting up a Generation for Failure by Jonathan Haidt and Greg Lukianoff

All-Time Keepers

Programming Books

A Tour of C++ by Bjarne Stroustrup
Python Essential Reference by David Beazley

Computer Systems

Designing Data-Intensive Applications by Martin Kleppmann
Operating Systems: Three Easy Pieces by Remzi Arpaci-Dusseau and Andrea Arpaci-Dusseau

Biographies

The Prince of Mathematics: Carl Friedrich Gauss by M.B.W. Tent
The Strangest Man: The Hidden Life of Paul Dirac, Quantum Genius by Graham Farmelo

Rafid Al-Humaimidi

Member of Technical Staff @ Cohere