Software Engineer II - AI Backend - Distributed Systems

Mithlesh Upadhyay

I build scalable AI backend systems, RAG workflows, LLM integrations, and distributed product platforms with strong reliability, testing, and developer workflow practices.

Contact me Resume GitHub Hugging Face LinkedIn

4+ years backend, AI engineering, and software delivery experience

5+ workflows AI red teaming, triage, CVE, model access, and RAG systems

4 providers Gemini, Grok, Azure OpenAI, and LiteLLM integrations

3 test layers unit, integration, and functional test coverage

Role Fit

Areas where I can add value for product and platform teams.

AI Backend Engineering

Python, FastAPI, AsyncIO, REST APIs, service modules, configuration layers, validation paths, and production-grade backend workflows.

LLM and RAG Systems

Multi-provider LLM integrations, embeddings, FAISS, pgvector, OpenSearch, BM25 search, ingestion pipelines, and evaluation flows.

Distributed Product Platforms

Kafka, Redis, Celery, PostgreSQL, SQLAlchemy, Alembic, Docker Compose, CI/CD, auth, multi-tenancy, ACL, RBAC, JWT/OIDC, and object storage.

Full-time or Consulting

Available for AI backend roles, RAG implementation, LLM platform work, reliability improvements, test automation, and developer tooling.

Experience

Engineering work focused on reliability, AI workflows, and backend ownership.

Software Engineer II

Emulate AI (formerly Detoxio AI) - Dec 2024 to May 2026

Contributed across a distributed microservices ecosystem for AI-driven security products, improving reliability, scalability, API behavior, service initialization, validation flows, integrations, and developer workflows. Built backend functionality across AI red teaming, assessment lifecycle, model integrations, triage services, CVE discovery, RAG-enabled systems, SDK/CLI tooling, and CI/CD plus Docker Compose validation.

SDE-I (AI Engineer)

Propellyr Inc. - Jan 2023 to Aug 2024

Built Python and FastAPI backend services, REST APIs, asynchronous service layers, PostgreSQL persistence, SQLAlchemy data-access layers, and Alembic migrations for AI chatbot, retrieval, and model-serving workflows. Developed LangChain RAG pipelines with chunking, embeddings, FAISS retrieval, context assembly, prompt orchestration, Celery workers, Redis, Kafka event flows, Docker Compose, Pytest coverage, and LLaMA 2 QLoRA fine-tuning.

Python Developer

Foresiet - Sep 2022 to Dec 2022

Built Python crawler utilities for APK data extraction, metadata processing, and cybersecurity research workflows. Developed FastAPI services and REST endpoints, Node.js integration utilities, React-based internal UI components, validation, error handling, logging, and reusable automation pipelines for internal security analysis workflows.

Senior Technical Content Engineer

GeeksforGeeks - Sep 2017 to Dec 2021

Led technical content and course-development work across computer science and competitive-exam domains, including GATE CS, while contributing 7,000+ articles, 3,000+ improvements, and 9,000+ review contributions.

Selected Projects

Production-style projects that show backend depth and AI system design.

RAG - Backend - ACL

Agentic RAG

Production-style multi-tenant RAG backend with ACL-aware retrieval, BM25 fallback, ingestion pipelines, and evaluation workflows.

GitHub Repo

LLM Gateway - Provider Control

Skyclad

AI multi-provider system built without a LiteLLM-style framework to reduce cost, preserve provider control, and keep routing logic explicit.

GitHub Repo

FastAPI - LLM APIs - Docker

AI Backend

Production-style AI backend APIs with LLM integrations, async services, testing, and Dockerized setup.

GitHub Repo

Provider Access - Validation

Service Ecosystem and Provider Validation

Backend workflows for multi-provider AI access, model validation, service configuration, automated checks, and repeatable integration testing across AI product services.

AI Security - LLM Robustness

AI Red Teaming Labs

Robustness and safety testing workflows for LLM-based systems, assessment lifecycle support, triage workflows, and backend automation for AI security products.

Solana - Web - Trading UI

ChadWallet Web

Solana API-powered landing and trading page inspired by fomo.family, built with a responsive web experience.

GitHub Repo Live Demo

Skills

Backend-first engineering with practical AI product experience.

Backend

Python FastAPI AsyncIO SQLAlchemy Alembic REST APIs Microservices Auth Multi-tenancy ACL/RBAC JWT/OIDC

AI Systems

LLM Integrations RAG LangChain LangGraph LiteLLM Embeddings BM25 OpenSearch FAISS pgvector

Infrastructure

Kafka Redis Celery PostgreSQL MongoDB Docker Docker Compose GitHub Actions Object Storage

Engineering Quality

Pytest Unit Testing Integration Testing Functional Testing SDK/CLI Tooling Automation Scripts Documentation

Working Knowledge

Rust Go HTML CSS JavaScript TypeScript Node.js React.js Next.js Express.js

Engineering Beliefs

How I think about building software.

Good software should be simple to run.

AI products need tests, not just demos.

Backend systems should fail clearly.

The best debugging starts with the exact error.

Documentation is part of engineering.

Speed matters, but correctness matters more.

Publication

A Comprehensive Analysis of Sentiment Analysis

Published with IEEE in 2024.

View IEEE Paper

Contact

Hiring for AI backend, RAG, LLM integrations, or distributed systems work?

I am open to full-time roles and consulting engagements where backend ownership, AI systems, reliability, and practical product building matter. Email me at mithleshupadhyay515@gmail.com

Email Resume LinkedIn GitHub Hugging Face