Pavan’s story

Architecting Fyers Automate

Designed orchestration, infrastructure, scaling, and recovery for a distributed trading-automation platform.

  • Distributed Systems
  • Temporal
  • AWS Architecture
  • Fault Tolerance

Pavan Kumar Segu

Senior Software Engineer 1 at FYERS

FFYERS
CCoinSwitch
3+ years of experience

From their time as

F

Software Engineer 2

FYERS • 2025 - 2026

Overview

Pavan designed the backend architecture for Fyers Automate, a no-code trading-automation platform. He owned orchestration, AWS topology, communication, scaling, and failure recovery from design through launch.

The Story

Pavan worked on Fyers Automate as a distributed platform that sends trading tasks across multiple microservices. Temporal acted as the orchestration layer, assigning work to the appropriate workers and maintaining workflow history.

He defined the parameters that governed task assignment, poller behavior, update frequency, response time, and recovery after failures. He also designed the end-to-end AWS architecture, including service placement, networks, connections, and deployment scripts.

The first infrastructure plan placed each lightweight microservice on a separate instance. Pavan revisited that design and consolidated compatible services onto shared instances, retaining horizontal scaling through automated scripts while reducing unnecessary infrastructure.

Unexpected load later exposed task failures and poller constraints. He adjusted the system through vertical scaling and increased poller capacity so the orchestration layer could absorb concurrent work more reliably.

The result was a self-managed platform with greater operational control, automatic recovery behavior, and an infrastructure model shaped around throughput and cost.

Ownership Snapshot

Broad role

Backend engineer with end-to-end architecture ownership for Fyers Automate.

Goal

Build a reliable and scalable platform for orchestrating automated trading tasks.

Direct ownership

Designed Temporal behavior, AWS deployment, service communication, scaling, and recovery.

Team execution

Worked with frontend and backend contributors to connect the platform's services and ship it.