Anish Keesari

Anish Keesari

Software Engineer | Backend Systems, AI/LLMs & Infrastructure | Python, Go, FastAPI, Kubernetes

Hyderabad, India

keesarianish@gmail.com


About

I'm a Software Engineer with a strong interest in backend development, AI engineering, and cloud-native systems. Currently working as an AI Engineer Intern at Roots Infosoft, I focus on building AI-powered applications, backend services, and reliable workflows that integrate large language models with real-world functionality. Previously, I worked as a Technical Intern at BARC, contributing to backend API development, data-driven systems, and performance testing. My experience spans Python, Java, Spring Boot, FastAPI, REST APIs, databases, and cloud technologies, along with hands-on exploration of LLM orchestration, retrieval-augmented generation (RAG), tool calling, multi-agent systems, and Kubernetes. Beyond internships, I enjoy building software projects that explore system design, automation, distributed workflows, and practical problem-solving. I'm particularly interested in understanding how scalable, reliable software is designed, deployed, and maintained. I'm always looking to learn from other engineers, exchange ideas, collaborate on meaningful projects, and connect with teams building useful technology. I'm open to opportunities in backend engineering, software development, and AI-enabled systems where I can contribute, grow, and take on challenging engineering problems.

Skills
TypeScript
JavaScript
Python
Java
Go
Next.js
Node.js
Google Cloud
Docker
PyTorch
PostgreSQL
MongoDB
CI/CD
FastAPI
LangChain
Agentic AI
LLM Orchestration
Retrieval-Augmented Generation (RAG)
Tool Calling
Embeddings & Vector Search
REST APIs
Redis
Kubernetes
AWS
Distributed Systems
Microservices
WebSockets
GitHub Actions
Unit & Integration Testing
Experience

AI development engineer

Roots Infosoft Technology

Aug 2026 to Present
  • Developing a production-oriented AI voice-support agent using Python, FastAPI, and LangChain. Designed model-agnostic orchestration with hybrid intent routing and retrieval-augmented generation (RAG) over a vector-indexed knowledge base. Built a schema-validated tool-execution loop with call budgets, deduplication, and strict argument validation. Developed agent memory using structured and semantic retrieval backed by PostgreSQL and SQLite, while contributing to integrations, testing, and deployment workflows.

Technical intern

Bhabha atomic research centre

May 2025 to Jul 2025
  • Developed FastAPI REST services for a quantitative research sandbox serving over 730 days of NSE market data and supporting backtesting of three trading strategies. Optimized asynchronous database workflows to achieve sub-500 ms API response times under simulated loads of 1,000+ concurrent users. Implemented Pydantic-based data validation and unit tests.
Projects

Kernexys

Built a Kubernetes reconciler for ModelDeployment custom resources with drift correction and status conditions, tested against a real kube-apiserver/etcd environment. Developed a typed FastAPI model registry and Redis-backed asynchronous inference workers with idempotency, backpressure, timeouts, and crash recovery.

Go
Python
Kubernetes
FastAPI
Redis

Multi-Agent AI Debate Platform

Built a real-time multi-agent system that orchestrates Gemini and Mistral agents through WebSockets and finite-state turn scheduling. Added prompt and output sanitization and human interjection to support controlled, interactive AI conversations.

TypeScript
Hono
Svelte
WebSockets
Gemini
Mistral

Offline-First Multi-Tenant Inventory SaaS

Implemented an offline synchronization engine using vector-clock conflict resolution. Designed multi-tenant data isolation with PostgreSQL Row-Level Security and secured APIs with JWT authentication and role-based access control.

Next.js
Fastify
PostgreSQL
IndexedDB

Weavers Club Sarees – E-commerce Storefront

Built and run a live e-commerce store for a real saree business. Includes an admin panel and a WhatsApp ingestion pipeline: the seller sends photos, videos and a description in chat, and the system turns them into draft products automatically. Handles out-of-order messages with an inbox queue, dead-letter handling and versioned DB migrations.

Next.js
Supabase (PostgreSQL)
Vercel
WhatsApp webhooks
Education

Class 10

Atomic energy central school 4 · 85.5%

Aug 2019 to Jul 2020

Class 12

Pace junior high school · 71%

Jul 2021 to Jul 2022

Bachelor's of technology in Computer Science

Vellore institute of technology · 7.41

Jul 2022 to Jul 2026
Certifications

AWS Cloud Practitioner

Amazon Web Services

Jun 2026

AWS Certified Developer - Associate

Amazon Web Services

Jun 2026

AWS Academy Graduate - Cloud Foundations

Amazon Web Services

Oct 2024

AWS Academy Graduate - Cloud Architecting

Amazon Web Services

Oct 2024
Anish Keesari · DevFest Hyderabad 2026