Hi, I'm

Temi
Lajumoke

I'm a Staff Software Engineer & Machine Learning Engineer who has spent over a decade building distributed systems and ML infrastructure. Most of my recent work has been LLM inference and model serving: the production model runtimes, model hosting and deployment pipelines, and agentic orchestration behind Amazon's Titan and Nova foundation models, where I was a founding engineer in the AGI Lab. Before that I shipped a bit of everything across the stack, from web and mobile apps to backend services and cloud native infrastructure.
I currently work on AI-powered tools and infrastructure at NVIDIA, leveraging Large Language Models and agentic AI to automate engineering workflows at scale.

Beyond engineering, I compose relaxing fingerstyle guitar music and recharge in the great outdoors, whether it's kayaking on tranquil waters or snowboarding down groomed slopes.

Current Personal Research Focus

I'm fascinated by making AI more accessible and efficient. Currently, I'm focused on building and training small language models (SLMs) that empower AI agents to run anywhere, while maintaining the power of their larger cousins. My work explores orchestration and observability strategies for agentic systems using these SLMs.

Recent (published) work include a GPT-inspired decoder-only model and a ground-up recreation of the original transformer architecture from the "Attention is All You Need" paper.

Skills

Languages

Core

TypeScriptJavaScriptPythonJavaC#HTML5CSS3SQL

Additional

SwiftGoRust

Software Frameworks

Node.jsExpressFastifySpring BootSpring CloudASP.NET CoreReactReact NativeNext.jsVueAngularKnockout.jsiOSJQueryReduxApolloModel Context Protocol (MCP)

API Specs

RESTGraphQLRPC / gRPC

Data stores and Queues

MSSQLPostgreSQLMongoDBMySQLFirebaseRedisRabbitMQKafkaSQSSNSDynamoDBElastic Search

Cloud & Orchestration

AWS (EC2, ECS, Lambda, Cloudwatch, Cloud Formation, CDK, VPC, S3, etc.)DockerKubernetesPrometheusGrafana

AI/ ML

PyTorchPandasTensorFlowKerasOpenCVScikit-learnSpaCyNLTKTransformersVector Space ModelsRNNs (LSTM, GRU)CNNsGANsLLM Runtime OptimizationRAGLangGraphLangChainLlamaIndexAgentic Systems

Experience

2015 - Present
2026-Present

Staff Software Engineer

NVIDIA

I work on AI tools and infrastructure that use LLMs and agentic AI to automate engineering workflows. Still cooking something great here.

2022-2025

Software Engineer

Amazon AGI Lab

Founding engineer for both the AGI Lab and Alexa on LLM (now Alexa+). I tech led the first model hosting pipeline and C2PA for Nova Reel, built much of the tool call and action execution layer (MCP before MCP), and worked on Reel's long video generation.

Nova technical report
2021

Machine Learning Engineer

BU Spark

Topic modeling and classification of historical records from scanned, multilingual archives. I used BERT and XLM-RoBERTa and improved identification of key themes by 11%.

2018-2020

Senior Software Engineer

Coursera (Senior Consultant via Andela)

I led a distributed team of eight engineers across three countries and was tech lead for accessibility compliance. I rewrote the in-browser code editor and the evaluator that runs learners' code, cutting bundle size for some services by up to 50%, and built the partner tools for creating specializations and customizing certificates.

2017-2018

Software Development Learning Facilitator

Andela

I facilitated coding classes for new software engineers and built the curriculum. I also mentored engineers and assessed candidates.

2015-2017

Software Developer

Woodstock

Smart home and home automation. I programmed control systems, built the apps people used to run their homes, integrated third-party devices into one system, and built custom IoT devices.

© 2014 - 2026 | Temi Lajumoke. All rights reserved.