UTC−4Start a conversation
Available for forward-deployed engineering

West Lafayette · Systems / AI / Cloud

Amogh AryaMunipalle

I build systems that have to work outside the demo.

Forward-deployed engineer working across edge AI, cloud infrastructure, and the last difficult mile between a prototype and a system people can trust.

View the thesis system Talk to me

100K+

Production requests / day

90%+

Edge vision precision

40%

Release cycle reduction

3.89

Purdue graduate GPA

EDGE · CLOUD · FIELD
AM / SYSTEMS2026
Amogh Arya Munipalle in front of Purdue's Bell Tower
SHIP / MEASURE / ITERATEFIELD
SYSTEMS
01 EDGE AI
02 CLOUD
03 DIGITAL TWIN

Flagship work · Purdue University · 2026

Thesis approved

A digital twin that listens, sees, and reasons at the edge.

A real-time safety system combining thermal and RGB vision, acoustic event detection, spatial tracking, and a live digital twin on Raspberry Pi hardware.

Try Thesis ER Read the source

I also open-sourced the formatter I built along the way: 89 deterministic Purdue checks, accessibility review, and zero file storage.

PHYSICAL SPACE
RGB
IR
AUDIO
EDGE FUSION
DIGITAL TWINLIVELATENCY · TRACKING · DECISION
90%+vision precision
93%audio accuracy
4fused signals
1edge system

Selected systems

Built to leave the slide deck.

Products, infrastructure, and applied AI that crossed the line into working systems.

Distributed Kubernetes Microservice Pipeline01SHIPPED
Feb 2025Purdue University, West Lafayette

Distributed Kubernetes Microservice Pipeline

YAML, Virtual Machines, Kubernetes, Docker, System Networking

  • Deployed a distributed system on a remote VM using Vagrant + VirtualBox with two Kubernetes pods running containerized services.
  • Enabled secure external access with NGrok tunnels and a Streamlit UI, forwarding results across pods via webhooks.
  • Automated container builds and rollouts with GitHub Actions and YAML workflows.
Iris-Based Multi-Disease Classification System02SHIPPED
Dec 2023Vellore Institute of Technology

Iris-Based Multi-Disease Classification System

MLOps, Docker, Computer Vision, APIs, MongoDB

  • Built a deep-learning pipeline to classify seven ophthalmic diseases using 50,000+ iris images, improving accuracy by ~18%.
  • Containerized the ML lifecycle to enable reproducible experiments and reduce environment issues by 95%.
  • Designed microservices for asynchronous ingestion, preprocessing, and inference with a FastAPI backend.
AI-Powered Stress & Productivity Analytics (I2safe)03SHIPPED
Sep 2024 - Nov 2024Purdue University

AI-Powered Stress & Productivity Analytics (I2safe)

Python, FastAPI, Docker, scikit-learn, pandas, NumPy, cvxEDA, Matplotlib

  • Engineered an end-to-end ML pipeline ingesting 500K+ physiological sensor readings — EDA, heart rate, and skin temperature — with automated regex-based cleaning and pandas transformations.
  • Applied cvxEDA decomposition and ensemble ML models (Random Forest, XGBoost) to classify stress and emotional states in real time with over 90% reliability.
  • Fused ambient environmental signals (temperature, humidity, CO₂) with biometric data, boosting productivity prediction accuracy by 15%.
Multi-Cloud DLP & Security Automation Platform04SHIPPED
Oct 2024 - Dec 2024Purdue University

Multi-Cloud DLP & Security Automation Platform

FastAPI, MongoDB, AWS S3, GCP Storage, Firebase, Streamlit, Mailjet API, Regex DLP Engine

  • Architected a multi-cloud security platform that continuously scans AWS S3 buckets and GCP Storage objects for misconfigurations, open permissions, and sensitive data exposure.
  • Built a custom regex-based DLP engine detecting PII, API keys, and secrets across 100+ cloud resources with 97% pattern-match accuracy.
  • Designed an automated alerting pipeline using Mailjet API — security violations trigger instant email reports to stakeholders, reducing incident response time by 20%.
Real-Time Payment & CRM Sync Engine05SHIPPED
Feb 2024Vellore Institute of Technology

Real-Time Payment & CRM Sync Engine

FastAPI, Stripe API, Salesforce API, Celery, RabbitMQ, Ngrok, Docker, SQLite, MySQL

  • Built a real-time event-driven microservice that syncs payment transactions from Stripe and customer records from Salesforce into a unified local database via webhook listeners.
  • Implemented asynchronous task processing with Celery + RabbitMQ, handling bursty webhook payloads without data loss or duplicate processing.
  • Exposed secure tunnel endpoints via Ngrok for local development, enabling end-to-end testing of Stripe checkout flows and Salesforce contact updates in real time.
Secure Blogger — Role-Based Publishing Platform06SHIPPED
Apr 2023Vellore Institute of Technology

Secure Blogger — Role-Based Publishing Platform

Python, FastAPI, JWT, OAuth 2.0, Docker, GitHub Actions, PostgreSQL

  • Designed and built a full-stack blogging API with granular role-based access control — admin, editor, and reader tiers — enforced via JWT tokens and OAuth 2.0 flows.
  • Implemented secure CRUD endpoints with input validation, rate limiting, and request signing to defend against injection and replay attacks.
  • Set up a CI/CD pipeline with GitHub Actions that runs linting, unit tests, and Docker image builds on every push, achieving zero-downtime deployments.

Field log

Experience

Roles where the work moved between customers, infrastructure, and deployed software.

01

Founding Software Engineer

Freight Pins

Remote, VirginiaDec 2023 - Dec 2025
  • Collaborated directly with clients to understand requirements and develop MVPs, architecting serverless cloud infrastructure processing 100K+ daily requests using AWS Lambda, DynamoDB, S3, SNS/SQS, SES, CloudWatch, and JWT encryption with 2FA.
  • Automated the end-to-end client onboarding process with a Node.js web automation framework for contract generation and integration, reducing manual effort by 60%.
  • Used Terraform to create staging environments and automate VPC provisioning, paired with CI/CD pipelines via Docker and GitHub Actions for zero-downtime deployments — cutting release cycles by 40%.
  • Led cross-functional initiatives and onboarded three developers; introduced performance monitoring with CloudWatch and DataDog to improve system reliability by 25%.
02

Research Assistant · AI Developer (Part-time)

Purdue Civil Engineering

Purdue University, West LafayetteMar 2025 - Jun 2025
  • Built a real-time shooter detection system on Raspberry Pi using YOLOv8 with thermal-RGB fusion, achieving ~90% precision and 86% recall.
  • Engineered an audio classification model using Mel-spectrograms, PANN embeddings, and CNNs to detect gunshots and screams with 93% accuracy.
  • Integrated visual, auditory, and thermal AI models into a Unity-based digital twin to classify threat levels and trigger automated alerts.
  • Collaborated with external stakeholders to understand project requirements, align deliverables with research timelines, and ensure on-time delivery of all milestones.
03

Research Assistant · Cyber Security Developer (Part-time)

Purdue Nuclear Engineering

Purdue University, West LafayetteAug 2024 - Jan 2025
  • Developed secure and encrypted communication protocols for two drones in flight, enabling real-time encrypted data exchange over long-range links using ESP32 microcontrollers with secure RNG and linear-algebra key generation.
  • Designed covert data-embedding protocols in microcontroller metadata and request timing for stealthy drone-to-drone communications.
  • Implemented encryption algorithms on ESP32/ESP8266 systems for IIoT data pipelines, adopted across Purdue labs to improve long-range secure communication.
  • Authored research documentation on lightweight cryptographic models for resource-constrained edge devices, contributing to ongoing nuclear security research.
04

DevOps Intern · Deployments Engineer (Part-time)

MysticLabs

RemoteJun 2023 - Aug 2023
  • Worked with Dell as a client, training custom LLMs on proprietary enterprise data using LangChain and vector databases (Pinecone, ChromaDB) with chunking strategies for accurate retrieval-augmented generation.
  • Built and deployed web crawlers using Apify to scrape and ingest large-scale datasets, feeding them into vector stores for LLM fine-tuning and semantic search pipelines.
  • Containerized microservices with Docker and implemented Celery with RabbitMQ to handle parallel AI inference workloads at scale.
  • Shipped product integrations to WhatsApp using the Twilio API and added four real-time webhooks for automated customer interactions.
05

Backend Developer Intern (Part-time)

SwiftSKU (Y Combinator)

RemoteJul 2022 - Aug 2025
  • Selected as a backend developer intern at a Y Combinator-backed startup, working extensively with AWS Step Functions to build 10+ state machines using Lambda for scalable serverless automation.
  • Designed and provisioned production VPCs with public/private subnets, NAT gateways, and security groups to isolate microservice workloads across AWS environments.
  • Implemented end-to-end CI/CD pipelines with GitHub Actions and Docker, automating build, test, and deployment workflows for rapid feature delivery.
  • Built AWS SNS/SQS pipelines for high-priority system alerts and event-driven architectures, ensuring reliable message processing across distributed services.

By the numbers

Measured outcomes

A few useful signals from the systems and teams behind the work.

100K+
Requests/Day

Architected serverless cloud infrastructure on AWS Lambda, DynamoDB, and SQS processing 100K+ daily requests with zero downtime.

40%
Faster Releases

Cut release cycles by 40% through automated CI/CD pipelines with Docker, GitHub Actions, and Terraform-managed staging environments.

90%+
AI Precision

Built real-time shooter detection on Raspberry Pi using YOLOv8 with thermal-RGB fusion, achieving ~90% precision and 86% recall on edge devices.

3
Research Assistantships

Held research assistantships at Purdue — encrypted drone comms for Nuclear Engineering, AI threat detection for Civil Engineering, and ML-based cone and construction safety detection models for the Indiana Department of Transportation.

Certified
Docker & Kubernetes

Completed Docker and Kubernetes: The Complete Guide (Udemy, May 2023) — applied containerization and orchestration across all production systems.

5+
Client Engagements

Embedded with clients as a forward deployed engineer — scoping MVPs, gathering requirements, and shipping production systems aligned to business outcomes.

Capability map

Tools I reach for

The stack changes. The job stays the same: understand the system, find the bottleneck, ship the fix.

Languages

PythonGoJavaScriptHTML/CSS

Systems & Virtualization

LinuxVirtualBoxVirtual MachinesVagrant

Frameworks

FastAPIFlaskDjangoGoGin

DevOps & Cloud

DockerKubernetesTerraformHerokuGitGitHub Actions

AWS & GCP

VPCLambdaS3EC2SNSECSIAMELBRDSRoute53

Databases

MongoDBMySQLRedisFirebaseAWS S3LangChain (vector)

Client Management

Stakeholder CommunicationRequirements GatheringMVP Scoping & DeliveryCross-functional CollaborationClient Retention Strategy

Operating principles

How I work

Close to the user, honest about constraints, and interested in systems that survive contact with reality.

I operate at the intersection of client needs and technical delivery, translating business goals into intelligent, cloud-native systems. I build end-to-end automated pipelines — from CI/CD and infrastructure-as-code to AI-powered workflows — that let teams ship faster and retain customers at scale.

As a forward deployed engineer, I embed with stakeholders and engineering teams to deliver outcomes fast. My work spans serverless architecture, DevOps automation, and applied AI — including building AI bots for real-time audio classification, voice-driven interfaces, and intelligent alerting systems that turn raw signals into actionable insights.

Current signal

Open to difficult, useful engineering work.

Forward-deployed roles, edge systems, cloud platforms, and teams that value people who can move between the customer and the code.

Academic record

Education

The formal work behind the field work.

Master's in Computer and Information Technology

Purdue University

West Lafayette, Indiana, USAAug 2024 - 2026

GPA: 3.89/4 · Master's thesis approved

Bachelor's in Electrical and Electronics Engineering

Vellore Institute of Technology

Vellore, IndiaMay 2020 - May 2024

GPA: 8.51/10

Open channel

Let’s build something real.

If the problem touches customers, infrastructure, or edge AI, I want to hear about it.

Transmission open

Bring me the problem that does not fit neatly in a job description.

I am most useful where software meets a real environment: a customer, a field device, a deadline, or a system with consequences.

Connect on LinkedInView GitHub

Location

West Lafayette, Indiana

LinkedInGitHub