Ayush Ranjan Roy

Ayush Ranjan Roy

AI Engineer at Accenture. Multi-agent systems, retrieval, and the platform work underneath them.

I contribute agent and retrieval components on enterprise Order-to-Cash programs, including Dispute & Deduction. Previously I was a platform engineer on Accenture’s MxDR / Adaptive MxDR product on AWS.

Claude Certified Architect — Foundations · Anthropic, May 2026

Experience

  1. Aug 2025 — Present

    AI Engineer · Accenture

    GenAI & Data

    • Multi-agent and workflow components for enterprise Order-to-Cash / Dispute & Deduction programs.
    • Multi-model inference (Claude, GPT, Azure) with routing, telemetry, and retrieval grounding.
    • Governed tool access and agent observability with OpenTelemetry and Datadog.
    • Architecture notes and enablement for engineers joining the program.
    LangGraphClaudeAzureFastAPIPostgreSQLOpenTelemetry

  2. Aug 2023 — Aug 2025

    Platform Engineer · Accenture

    Security · MxDR / Adaptive MxDR

    • AWS platform work modernizing MxDR → Adaptive MxDR: Terraform, CI/CD, production support across shared services and multi-account environments.
    • Redis-backed shared metadata services; multi-environment ops; Lambda runtime upgrades (Python 3.8 → 3.12).
    • Secrets Manager and IAM automation; service and container hardening.
    • Telemetry into enterprise SIEMs (mTLS, OpenTelemetry); blue-green deploys on live upgrades.
    AWSTerraformRedisSecrets ManagerIAMCI/CD

Selected work

  • Order-to-Cash / Dispute & Deduction agents

    Enterprise

    Multi-agent workflow components, durable state, and human review gates on a multi-engineer Accenture GenAI program.

    LangGraphRAGFastAPIPostgreSQL

  • Multi-model inference

    Enterprise

    Async inference across Claude, GPT, and Azure with routing and cost/latency telemetry.

    ClaudeGPTAzure AI

  • MxDR → Adaptive MxDR on AWS

    Enterprise

    Platform modernization for Accenture’s security product: Terraform, shared services, CI/CD, secrets, multi-env ops.

    AWSTerraformRedisCI/CD

  • HelixOps

    Open source

    Incident IDE: multi-agent RCA over runbooks, then human-approved remediation. Mock-safe by default.

    FastAPIRAGReact
    Source

  • RAGGym

    Open source

    Retrieval practice platform with swappable models and stores, self-correction, and an evaluation harness.

    RAGLangGraphStreamlit
    DemoSource

About

I’m an AI engineer at Accenture (GenAI & Data). Day to day I work on multi-agent workflows, inference routing, tool access, and human review steps for Order-to-Cash / Dispute & Deduction delivery programs — as a contributor on multi-engineer teams.

From 2023–2025 I was a platform engineer on Accenture Security’s MxDR → Adaptive MxDR product: Terraform, shared services, CI/CD, secrets, and telemetry on AWS across multi-account environments.

Outside work I maintain HelixOps and RAGGym. Based in Gurugram. Open to senior AI roles.

Education

B.Tech, Electronics and Communication Engineering · VIT Chennai, 2019–2023

Artificial IntelligenceEssentials of Machine LearningCryptography and Network SecurityInformation Security Analysis and AuditData Structures and AlgorithmsOperating Systems

AI & systems

Multi-agent systemsRAGLangGraphClaudeGPTAzure AIFastAPIPostgreSQL

Platform

AWSTerraformIAMCI/CDOpenTelemetryDockerRedis

Contact

Best reached by email. Open to senior AI roles.