// Software Engineer & Graduate Researcher

Motheo
Treasure Puso

Distributed Systems · Cloud Engineering · Applied AI

I build systems that perform at scale — distributed pipelines processing millions of events per hour, fault-tolerant microservices, and AI-powered applications that solve real problems. Currently pursuing my MS in Computer Science at UT Arlington, deepening my focus on Cloud Computing and Big Data. My dual background in CS and Finance from Michigan State shapes how I think: not just what to build, but why it matters and what it costs to get it wrong.

View Projects ↓ Download Résumé Let's Connect
Motheo Treasure Puso
Motheo Treasure Puso
Arlington, TX
~/motheo-puso — zsh
❯ cat profile.json
{
  "name": "Motheo Treasure Puso",
  "role": "Backend / Distributed Systems",
  "location": "Arlington, TX",
  "focus": [
    "Cloud Engineering",
    "Big Data & Streaming",
    "Applied AI / ML"
  ],
  "education": "MS CS @ UTA",
  "cert": "AWS Solutions Architect",
  "status": "open to opportunities"
}
❯

// 02 — Technology

Tech Stack

Big Data & Streaming
Apache Spark Kafka Hadoop HDFS MapReduce Spark SQL MLlib Structured Streaming
Cloud & DevOps
AWS Docker Kubernetes Terraform EC2 EKS S3 RDS CloudWatch CI/CD
Languages
Python Java SQL JavaScript Bash
Backend & AI
Flask LangChain gRPC Django RAG Vector Embeddings Anthropic Claude
Databases
PostgreSQL Redis Raft Consensus ACID
Methodologies
Agile / Scrum Microservices Distributed Systems Test Automation Load Testing Grafana

// 03 — Selected Work

Projects & Builds

001 — Featured

Real-Time Anomaly Detection Pipeline on Distributed Data

Apache Spark Kafka Hadoop MLlib Isolation Forest MapReduce HDFS

A production-grade streaming pipeline built on a multi-node Hadoop cluster, ingesting and processing over 10 million events per hour with sub-500ms anomaly detection latency. The system combines Spark Structured Streaming with Kafka for real-time ingestion, custom MapReduce workflows for feature engineering over 50GB+ log datasets, and a fully integrated Isolation Forest model via Spark MLlib for unsupervised anomaly detection at scale.

68% reduction in preprocessing runtime vs. single-node execution via distributed MapReduce Sub-500ms real-time detection latency across continuous 10M+ events/hour throughput End-to-end pipeline: Kafka → Spark Streaming → HDFS → ML inference
⚙ In Progress — GitHub Coming Soon
002

AI Support Chatbot — Ally Financial

An investor relations chatbot built during a university-industry collaboration with Ally Financial, enabling shareholders to access timely and relevant company information through a conversational AI interface. Designed for production reliability with enterprise-grade authentication and cloud-native deployment.

30% boost in user engagement through context-aware AI responses 25% reduction in query latency via prompt engineering and vector embeddings Amazon Cognito auth + EKS deployment via GitLab CI/CD pipeline
Anthropic Claude 3 LangChain AWS EKS Flask RAG Cognito S3
View Project ↗
003

Distributed Library Management System

A fault-tolerant microservices architecture demonstrating deep expertise in distributed systems design. Implements the Raft consensus algorithm from scratch for leader election and log replication, ensuring ACID properties across independently deployed services. Validated under simulated network partitions and cascading node failures to meet a 99.999% availability target.

Raft consensus: leader election, log replication, and graceful failure recovery Handles 3× peak load with zero data loss under network partition scenarios Fully containerized with Docker; infrastructure provisioned via Terraform
gRPC Docker Terraform PostgreSQL Raft Microservices
View Project ↗
004

ASMSU Student Government Website

Designed and built the public-facing website for the Associated Students of Michigan State University — a platform serving over 30,000 students. Focused on accessibility, performance, and maintainability, with a responsive layout and dynamic content that keeps students informed in real time.

Dynamic content rendering and real-time API integrations for campus updates Responsive design serving 30,000+ students across MSU's student body Maintained with Git version control and continuous deployment workflows
React.js Node.js REST APIs CSS CI/CD
Visit Site ↗

// 04 — Writing

Blog & Research Notes

Feb 2026
Big Data

Building a Streaming Pipeline: Lessons from 10M Events/Hour

What nobody tells you about Spark Structured Streaming in production — backpressure, watermarking gotchas, and how we cut preprocessing time by 68% with custom MapReduce workflows.

Read More →
Jan 2026
Distributed Systems

Raft Consensus in Practice: What the Paper Doesn't Explain

Implementing Raft from scratch for our distributed library system exposed edge cases in leader election and log replication that the original paper glosses over. Here's what I learned.

Read More →
Coming Soon
Cloud Engineering

Terraform at Scale: Managing AWS Infrastructure Without Losing Your Mind

A practical guide to module organization, state management, and the lessons learned provisioning multi-region infrastructure on AWS.

Coming Soon
Dec 2025
Applied AI

RAG Pipelines That Actually Work: Embeddings, Chunking & Context

After building two RAG-powered systems in production, here's my honest breakdown of what chunking strategies work, which embedding models to reach for, and where most demos fall short.

Read More →
Coming Soon
Research

Anomaly Detection at Stream Velocity: Isolation Forest vs. Statistical Methods

A deep dive into the tradeoffs between classical statistical anomaly detection and ML-based approaches when you're constrained to sub-500ms latency budgets.

Coming Soon
Nov 2025
How-To

From Zero to Kafka: Setting Up a Multi-Broker Cluster on AWS

A step-by-step walkthrough of deploying a production-grade Kafka cluster on EC2 — broker configuration, partition strategy, and monitoring with CloudWatch.

Read More →

// 05 — Experience

Work & Education

UT Arlington
Aug 2025 — Present
MS Computer Science — Cloud & ML

Focused on Distributed Systems, Machine Learning, Cloud Computing & Big Data, and Software Testing. Active research in streaming data architectures and real-time ML inference pipelines.

Distributed SystemsMachine LearningCloud & Big Data
Ally Financial
Jan 2024 — May 2024
Software Engineer Intern — University-Industry Collab

Built and shipped an AI support bot using Flask, LangChain, and AWS that drove a 30% boost in user engagement. Reduced LLM query latency 25% through prompt engineering and embedding optimization. Deployed on AWS EKS with a RESTful API powering a React frontend.

FlaskLangChainAWS EKSReact
Altair Engineering
May 2022 — Dec 2022
Software Engineer Intern

Automated data validation pipelines with Python (Pandas, NumPy), reducing processing errors by 22% and improving pipeline efficiency by 35%. Developed PyTest suites to validate software performance and identify system bottlenecks.

PythonPandasPyTestMonarch
MSU Student Gov.
Mar 2022 — Aug 2024
Director of Technology — ASMSU

Led technology strategy for a platform serving 30,000+ students. Built the MSU Spotlight news app in React Native with 5,000+ downloads. Implemented Grafana monitoring dashboards and maintained CI/CD workflows across the university's student-facing infrastructure.

React NativePythonGrafanaCI/CD
Michigan State Univ.
Aug 2019 — Aug 2024
BS Computer Science & BA Finance — Double Degree

Graduated with dual degrees combining rigorous CS foundations with financial theory — a pairing that informs both the technical architecture and business logic of everything I build. National Top Achievers Scholarship recipient (Botswana).

AWS Certified Solutions ArchitectNational Top Achiever Scholar
// 06 — Let's Work Together

Open to Ambitious
Opportunities

If you're building something at the intersection of cloud infrastructure, distributed systems, or AI — I'd love to hear about it. I'm actively seeking full-time roles and research collaborations for 2026.

Send an Email ↗ LinkedIn ↗ GitHub ↗ Résumé PDF ↗