Software Engineer · New York

Bryan Mejia

Software Engineer interested in agentic AI, full-stack development, and cloud infrastructure.

PROJECTS CONTACT ME

Bryan.

Who I Am

I'm a Computer Science graduate from CUNY Queens College and currently a Forward Deployed Engineer at Augenta AI. I lead an engineering team to develop an enterprise grade contextual memory platform deployed in production for multiple enterprise customers. I've enjoyed building and shipping agentic AI systems as well as working with Microsoft Azure to learn infrastructure.

What I Do

My focus is developing agentic AI systems; agent harness integration, lifecycle-hook context injection, tool calling, and RAG pipelines. I've worked with data pipelines that capture agent trajectories across harnesses and user-agent interactions to unlock the power of contextual memory in agentic systems.

Beyond the Code

I've been fortunate enough to experience working in other fields, such as fitness and operations. I was a personal trainer and fitness coach for 3 years, and an operations assistant for 4 years. Weightlifting is a big part of my life, which led me to rediscover myself through nature and the outdoors. I enjoy hiking, camping, and exploring new places. I also enjoy reading, especially books on philosophy, psychology, and technology.

Relevant Coursework

Data Structures & Algorithms
Design and Analysis of Algorithms
Machine Learning
Applied Data Science
Software Engineering
OOP in Java & C++
Operating Systems
Database Systems
Internet & Web Technologies
Computer Architecture
Quantum Computing

Off the map

Places that I've Visited!

The toolkit

What I work with

AI & Agentic Systems
Languages & Frameworks
Cloud & Infrastructure

A slowly turning constellation of everything I build with; hover any skill to read it, or hover a group to focus it.

The road so far

Experience

01 / 05

keep scrolling to flip through

Experience Jan 2026 - present

AI Engineer

Forward Deployed Engineer

Augenta AI · New York

Lead infrastructure engineer for an Azure cloud-native contextual memory platform deployed in production for enterprise customers.
Developed a plugin for multiple agent harnesses capturing agent trajectories and user-agent interactions across all coding sessions, streaming telemetry through an OpenTelemetry → Event Hubs producer/consumer buffer into a downstream memory engine.
Created multi-agentic workflows for testing and deployment. Implemented a robust CI/CD pipeline to streamline the deployment process.
Engineered lifecycle-hook context injection that reused shared context across users, agents, and sessions; cutting token usage by 24% and improving agent recall by 33%.

AzureKubernetesOpenTelemetryEvent Hubs

Experience Mar - May 2026

Google SWE Fellow

BASTA

Competitively selected from a national applicant pool for a Google-backed software engineering fellowship; mentored directly by a Google Cloud Engineer.
Completed 18 of 20 technical interviews in system design, data structures, and algorithms problems across structured curriculum.
Cooperated with other fellows to share knowledge and best practices, attending BASTA hosted events.

System designDS & AMentorship

Experience Jun 2022 - Apr 2026

Operations Assistant

MQ Steel Corp · New York

Optimized cross-functional workflows across 25+ active client accounts, reducing average turnaround time by 15%.
Maintained client database with 100% accuracy in financial record-keeping, processing $500K+ in annual transactions.
Collaborated with management to identify improvement opportunities and implement solutions, navigated 5 boroughs to meet with clients to discuss planning.

OperationsProcess designData accuracy

Leadership Feb 2024 - Jan 2025

Treasurer / Event Coordinator

SIAM · Queens College

Allocated $7,500+ across professional development and hackathon sponsorships serving 200+ members. Ran weekly Python workshops and math competitions, driving a 70% increase in active participation over two semesters.
Managed a team of 10+ volunteers to plan and execute events, ensuring smooth coordination and execution.
Organized and led 5+ successful events, including conferences, enhancing community engagement.

BudgetingWorkshopsCommunity

Built and shipped

Projects

01 / 03

keep scrolling to flip through

Project Jan 2026

Security Sentinel

AI-driven code auditor

Orchestrated a multi-agent system for static analysis of large-scale Python repositories, using tool calling and structured LLM outputs to identify logic vulnerabilities with 85%+ detection accuracy.
Implemented Agentic RAG to inject project-specific context into LLM prompts, improving detection precision by 30% over a zero-shot baseline.
Containerized the full pipeline with Docker for reproducible analysis runs.

TypeScriptClaude APIMulti-agent RAGDocker

Project Oct 2025

Polly Debate AI

Real-time debate coach

Integrated DeepFace and OpenCV for real-time emotion detection, processing video frames at 30+ FPS during live debates.
Built multimodal scoring combining facial expression data with audio tone analysis, improving user debate performance metrics by 56%.
Streamed live analysis over WebSockets, with Google Gemini powering feedback generation.

PythonDeepFaceOpenCVGemini API

Project Nov 2024

HeartScope

Heart disease prediction

Built a full-stack health prediction system achieving 88% accuracy with Random Forest and sub-second risk assessment.
Optimized the ML pipeline with modal imputation and categorical encoding, reducing prediction errors by 12% on incomplete inputs.
Served predictions through a Java Spring backend with an interactive JavaScript front end.

JavaSpring BootJavaScriptRandom Forest

"The best way to predict the future is to invent it" - Alan Kay

Looking forward to new opportunities, new connections and new challenges!

Say hello