Skip to content
BACKEND & AI INFRASTRUCTURE·HYDERABAD, INDIAAVAILABLE · FULL-TIME & FREELANCE

Anyone can call an LLM. I build the system that runs it.

Senior engineer for the backend behind AI products: identity, data pipelines, realtime agents, and the deployment that keeps them alive.

Work with me →See the work ↓Resume ⇩

scroll. scattered signals wire themselves into a network

70%fewer auth bugs
200K+records in ETL
75%fewer agent tokens
3PRs merged into Typesense (26K★)

What I do

problems I take on

Backend systems

APIs and data layers that hold under real load.

FastAPIPostgreSQLRedisCelery

AI infrastructure

LLM features wired into production, not demos.

MCPRAGpgvectorRealtime agents

Identity & access

OAuth 2.0 / OIDC providers and SSO, built from scratch.

OAuth 2.0OIDCJWTPKCE

Data & deployment

Pipelines, search, and the Kubernetes that runs them.

AirflowPySparkTypesenseGKE

Selected work

five systems, one number each
3 → 1auth systems, 70% fewer bugs

One identity platform for five products

OAuth 2.0 / OIDC provider, 29 endpoints, SSO on day one.

10xScale.ai
FastAPIPostgreSQLRedisEdDSA
0hard crashes in realtime voice

AI interviews that stay up mid-call

Barge-in handling, worker prewarming, graceful time limits.

interviewpilot
Realtime voiceElectronML proctoring
10×faster search over 200K+ records

Idempotent ETL feeding hybrid search

Airflow and PySpark into pgvector and Typesense, on GKE.

careerpilot
AirflowPySparkpgvectorKEDA

Experience

where the work happened
  1. 2025 → now

    Senior Software Engineer

    10xScale.ai

    Identity platform, realtime agent infrastructure, GKE deployment.

  2. 2023 → 2025

    Software Engineer

    10xScale.ai

    ETL at scale, hybrid search, upstream C++ into Typesense.

  3. Education

    B.Sc. Computer Science

    Government Arts College, Coimbatore

Stack

what I build with
languages
PythonTypeScriptJavaC++SQL
backend
FastAPIDjangoSpring BootCeleryPydantic
auth
OAuth 2.0OIDCJWT (EdDSA)PKCESSORBAC
ai
MCPLLM APIsRAGpgvectortree-sitter
data
AirflowPySparkPostgreSQLRedisTypesense
infra
DockerKubernetesGKEAWSTerraformCI/CD
frontend
React 19TanStack QueryElectron
THE OVERLAPBackend depthAI integrationOpen sourceme

Most engineers have one. My systems ship because I work in the overlap.

WORK WITH ME

Bring me the problem. I'll build the system.

Two engagement models, one standard of work. Choose the fit for your team.

Full-time role

Senior backend, platform, or AI infrastructure. Long-term ownership.

  • Own auth, data, or agent infrastructure end to end
  • Teams shipping AI features into production
  • Remote, or on site in Hyderabad
level
senior
start
30 days notice
where
remote / Hyderabad
Discuss a role

Contract & consulting

A defined problem, solved and deployed on your infrastructure.

  • OAuth 2.0 / OIDC identity providers and SSO
  • MCP servers, RAG pipelines, agent backends
  • FastAPI services shipped on Kubernetes
scope
fixed or retainer
start
within 2 weeks
hours
IST, overlaps EU + US
Scope the problem
HOW AN ENGAGEMENT RUNS
  1. 01
    Define the problemA 30 minute call. You get a written plan and a fixed quote.
  2. 02
    Build in the openWeekly demos, a repo you own from day one.
  3. 03
    Deploy and hand overShipped to your infra, with docs and a runbook.

Let's build systems that hold.

Tell me what's breaking, or what you're building. I reply within a day.

© 2026 Mutharasu A · Hyderabad, INchaos → system · measured