Portfolio — AI • Cloud • Renewable Energy

About Me

Computer Science student pursuing an M.S. in AI and Machine Learning at Colorado State University Global (GPA: 3.82), with a background that sits at the intersection of operations, cloud engineering, and applied AI.

I don't just study AI, I build with it. From autonomous email agents and RAG-powered chatbots to multi-node LangGraph workflows and AWS-hosted production systems, I've spent the last year turning AI concepts into working tools that solve real business problems.

My experience spans five years managing complex technical operations at scale at Sunrun, a software engineering internship at an AI startup, and a current HR AI Fellowship at IMA Financial Group, where I'm embedded inside the HR leadership team building AI prototypes that automate manual workflows and surface workforce insights faster than any manual process could.

I specialize in the layer where AI meets real operations, finding where processes are broken or manual, building something that fixes it, and measuring the result to prove it worked.

Skills

Programming

Python JavaScript SQL

Cloud (AWS)

EC2LambdaS3 RDSDynamoDB VPCLoad Balancer API GatewayIAM CloudWatch

Tools

SalesforceTableau GitJiraVS Code

Technical Competencies

Root Cause Analysis Incident Escalation Management Telemetry & Log Analysis

Education

Colorado State University – Global

Denver, Colorado

B.S. & M.S. in Computer Science

Expected May 2027

Concentrations: Artificial Intelligence and Machine Learning
GPA: 3.82 / 4.00
Relevant Coursework: Data Structures & Algorithms, Objects & Design, Computer Organization & Programming, Combinatorics, Machine Learning, Artificial Intelligence, Object-Oriented Programming, Statistics & Applications

Certifications

AWS Certified Cloud Practitioner (CLF-C02)

March 2026

Covers S3, EC2, Lambda, RDS, VPC, IAM, and CloudFront
Validates understanding of AWS core services, cloud architecture, pricing, and security

View Certification

Projects

Customer Support Email Agent

Built an end-to-end AI email agent using a 9-node LangGraph workflow that classifies intent, runs FAISS semantic search against a knowledge base, and generates GPT-4 responses reducing simulated response time from 30 minutes to 3 seconds while routing complaints and low-confidence cases to human review. Architected a production-ready full-stack system across FastAPI, vanilla JS, and SQLAlchemy with six decoupled service modules, graceful fallback handling, and comprehensive logging automating approximately 80% of routine support cases end-to-end.

FastAPILangGraph FAISSGPT-4 PythonSQLAlchemy JavaScriptAI Agents

View on GitHub

AWS Serverless Task Management API

Built a full stack serverless task scheduler using React and TypeScript with an AWS backend powered by API Gateway, Lambda, and DynamoDB, implementing REST endpoints for full CRUD operations and scalable data storage. Diagnosed and resolved cross-service failures by fixing IAM permission issues, correcting a broken DynamoDB update pattern, and ensuring proper request payload handling.

API GatewayLambda DynamoDBReact TypeScriptIAM CloudWatch

View on GitHub

Gmail AI Assistant Agent

Built an autonomous AI agent in Python that integrates with Gmail via Google's REST API and OAuth 2.0 authentication, using Claude AI (Anthropic) to automatically summarize, categorize, and score urgency of incoming emails. Architected across four decoupled modules using the Perceive, Decide, Act loop with automated Gmail API replies and real-time desktop notifications.

PythonGmail API OAuth 2.0Claude AI REST API

View on GitHub

Denver Broncos NFL IQ

Built a Denver Broncos NFL IQ system using Python and pandas to process 17,000+ rows of NFL statistics into a structured knowledge base. Integrated AWS Bedrock (Claude 3 Haiku) via boto3 to ground AI responses in real Broncos data, eliminating hallucinations.

Python (Programming Language) AWS BedrockClaude 3 Haiku Retrieval-Augmented Generation (RAG) Pandas (Software)

View on GitHub

Three-Tier AWS Web Application

Designed and deployed a production-grade three-tier web application on AWS from scratch, implementing enterprise cloud architecture patterns used across the industry. Built a custom VPC with 8 subnets across two availability zones, enforcing network segmentation between public, application, and database tiers. Deployed Python Flask application servers behind Gunicorn, fronted by Nginx reverse proxies and two Application Load Balancers — one public-facing, one internal. Configured EC2 Auto Scaling Groups with target tracking policies between 2 and 4 instances based on CPU utilization. Provisioned an encrypted Amazon RDS PostgreSQL 16 database in isolated private subnets with automated backups enabled.

AWS VPCEC2 RDS PostgreSQL Elastic Load Balancing Auto ScalingNAT Gateway Security GroupsPython FlaskNginxGunicorn AWS CLI

View on GitHub

AWS Static Portfolio Website Deployment

Designed and deployed a static portfolio website using S3 for storage and CloudFront for global content delivery. Configured Route 53 for custom domain routing and DNS management. Implemented HTTPS using AWS Certificate Manager and optimized delivery with CDN caching for low latency performance.

S3CloudFront Route 53ACM CDN

View Live Site

Explore My Work on GitHub

Browse my repositories, contributions, and open-source projects

github.com/Aesher7

Professional Experience

Applied AI Fellow

IMA Financial — Denver, Colorado

Jun 2026 – Present

Embedded within IMA's HR leadership team to identify and automate manual, repetitive HR workflows, building deployable AI prototypes using Python, FastAPI, and LangGraph that reduce operational overhead and free up HR capacity for higher-value work
Partnering with People Analytics and HR Operations to design AI-assisted tools including an Ask HR assistant and automated data pipelines, translating ambiguous business problems into structured, measurable solutions with quantifiable impact on team productivity

Service Technical Efficiency Support

Sunrun — Remote, Colorado

Sep 2020 – Feb 2026

Managed end-to-end customer success workflows in Salesforce CRM, translating complex technical faults into clear remediation plans and communicating deployment status and outcomes to engineering leads and client-facing stakeholders across thousands of distributed solar systems.
Monitored system behavior and tracked performance degradation across a large distributed fleet using real-time telemetry and log inspection diagnosing and resolving unexpected configuration drift before it impacted customers, reducing field dispatch costs by remotely remediating 25% of escalated cases.
Accelerated resolution turnaround for high-priority customer cases by translating loosely defined technical faults into clear problem statements and executable remediation plans, documenting findings and communicating outcomes to cross-functional stakeholders including engineering leads and client-facing teams.

Software Engineer Intern

Qookit AI — Remote, Colorado

May 2024 – Nov 2024

Reduced manual integration effort across AI and data processing workflows by developing modular frontend components in JavaScript and HTML and building cross-platform mobile applications in Flutter and Dart, connecting backend AI pipelines to production interfaces via Python.
Prototyped and validated AI-enabled features by integrating backend data processing pipelines with client-facing interfaces, contributing to a real delivery initiative focused on applied AI usage within a fast-moving startup environment.

Architecture Diagrams

AWS system architecture designs and cloud infrastructure diagrams