Backend Developer

Akshith Macharla

Building scalable backend systems and cloud infrastructure with ML expertise. Specializing in Spring Boot, Django, distributed systems, and AWS with a focus on performance optimization.

Get In Touch View My Work
2.8
Years Experience
10K+
Active Users Served
99.9%
Uptime Maintained
4.0
GPA at CSU Chico
Career Journey

Professional Experience

Nov 2025 – Present
California State University, Chico
Web Developer
  • Owned and shipped the Chico Love event website end-to-end, delivering the full user flow from landing pages to registration UX to backend integration.
  • Implemented the registration backend on Cloudflare Workers + D1 with schema/index design, validation, and idempotent email uniqueness
  • Built a repeatable Wrangler local/remote deploy + test pipeline so the team can reproduce, verify, and ship changes confidently.
Dec 2021 – Aug 2024 (2.8 yrs)
Cognizant Technology Solutions
Programmer Analyst - Django Developer
  • Built and maintained Java Spring Boot microservices for insurance claims & policy workflows (10,000+ active users, SLA backed); improved API responsiveness via DB query optimization and payload tuning validated in Dynatrace.
  • Owned AWS production support and releases: monitored via Dynatrace, triaged/resolved 100+ incidents (auth/login, registration, integrations), and sustained 99.9% uptime.
  • Improved delivery quality with JUnit/Mockito tests and documentation, reducing onboarding time for new engineers.
Jan 2023 – July 2023
Edurekha
Data Science & ML Intern
  • Cleaned and prepared ~1.3M rows in reproducible Pandas/SQL pipeline
  • Trained regularized logistic regression achieving AUC improvement from 0.71 to 0.84
  • Improved minority-class recall by 20 points through threshold tuning and probability calibration
Featured Work

Projects

From ML systems to distributed platforms, here are some highlights of my recent work

Hardware-Aware Training Time & Throughput Prediction

TensorFlow • Keras • NVIDIA A100 • CIFAR-10

ML system to predict training time per epoch and throughput for CNN configurations on the department’s NVIDIA A100.

  • Ran multiple CNN configs on A100; logged time/epoch, images/sec, parameter count, and accuracy via callbacks and nvidia-smi
  • Built a dense ANN regressor in TensorFlow/Keras and compared against scikit-learn linear regression baselines
  • Framed as early job time estimation / hardware-aware scheduling prototype (A100, AWS Trainium, AWS Inferentia)
View Project →

RAGops Copilot

LangChain • LangGraph • Gemini API • Qdrant • FastAPI • Next.js

Developer-focused RAG + agent copilot that answers questions about codebases/docs/issues with strict, auditable citations and can take actions.

  • Built a RAG + agent workflow with LangGraph tool-calling to open GitHub issues and generate PR checklists
  • Implemented hybrid dense + BM25 retrieval with metadata filters and reranking
  • Added observability for p95 latency per cost; fine-tuned a small open model with LoRA; deployed streaming inference with caching and traces
View Project →

Task Platform (Distributed Job Queue)

FastAPI • RabbitMQ • Postgres • Redis • Prometheus • Grafana • Docker

Dockerized async task platform with reliability guarantees, retries, and observability for end-to-end throughput.

  • Built at-least-once processing with Idempotency Key deduplication, retries, and DLQ handling
  • Load tested 2,000 tasks: throughput improved from 31.25 tasks/sec (1 worker) to 41.7 tasks/sec (3 workers), +33%
  • Added observability with Prometheus metrics and Grafana dashboards
View Project →

University Marketplace

Django • PostgreSQL • Docker • GCP • Bootstrap

Marketplace platform with a containerized backend and high-availability deployment on Google Cloud.

  • Packaged Django + PostgreSQL services in Docker; deployed across multiple GCP VMs behind an HTTPS Load Balancer for high availability
  • Used GCP Storage Buckets for secure image hosting and added a server_info endpoint for real-time environment/config visibility
  • Secured 100+ registered users within 24 hours of launch
View Project →
Technical Arsenal

Skills & Technologies

Languages

Java Python C++ TypeScript CUDA

ML & Data Science

PyTorch TensorFlow Keras scikit-learn Pandas NumPy

Backend & Web

Spring-Boot Django FastAPI React Next.js REST APIs

Cloud & Infrastructure

AWS GCP Docker PostgreSQL Redis CI/CD
Achievements

Key Milestones

🏆
2nd Most-Used AI Agent
7.2K interactions on Fetch.ai track @ Cal Hacks 12.0 with Chaos Reviewer
🚀
Rapid Product Launch
100+ users in first 24 hours on University Marketplace
⚡
API Performance Optimization
Reduced p95 latency ~250ms → ~200ms, +15% throughput via Redis & ORM tuning
Let's Connect

Get In Touch

Currently pursuing M.S. in Computer Science at CSU Chico. Open to opportunities in ML Engineering and Backend Development.

📧 Email Me 💼 LinkedIn 💻 GitHub 📱 (530) 965-7864