About the role

STFATG Consultancy Pvt. Ltd. (Standardise The Flow, Automate To Grow) builds and runs workflow automation for clients in India and abroad. You will own builds end to end - design, build, deploy, document and support - across n8n Cloud workflows running off ERP exports (TCS iON, export-only, no live integration), a Google Drive/Sheets backbone with Apps Script, and Python tools.

Role & responsibilities

- Design, build, run and maintain multi-branch n8n Cloud workflows end to end: webhooks, credentials, scheduling, error handling and retries, within Cloud constraints (no local filesystem, I/O via Google Drive, explicit Asia/Kolkata timezone).
- Build Python data pipelines on ERP exports, reconciliation logic, Streamlit tools and API scripts; all numeric computation lives in Python or Sheets/Excel, never inside an LLM step.
- Create and operate agentic AI modules on LLM APIs (Claude/OpenAI): prompt design, tool-calling, guardrails and logging; the agent drafts, classifies and explains, while deterministic code computes.
- Build client/NGO portals on Hostinger: front end plus auth, role-based access and file handling on the Supabase/PostgreSQL pattern.
- Host and operate applications on Hostinger VPS: SSH, Docker, Nginx, domains/DNS/SSL, GitHub Actions deployment, separate staging and production, service isolation.
- Automate Sheets, Drive and Gmail with Google Apps Script: triggers, custom functions, register and escalation flows on the Google Workspace backbone.
- Read, debug and integrate with Java-based systems where client stacks require it (not a primary build language).




- Document every workflow and deployment and verify it against client figures before go-live.

Working discipline expected

- Plan before execute: a numbered skeleton plan is agreed before any build starts, and a visual mockup precedes every workflow, portal or dashboard.
- ERP integration is export-based only; never attempt live hooks into client ERPs.
- Computation belongs in Python, Sheets or Excel; the LLM writes the verdict, never the number.
- Data integrity: label stated vs derived figures, and surface contradictions in source data rather than resolving them silently.
- Every workflow and deployment is documented and verified against client figures before go-live.

Preferred candidate profile

- Python: expert, primary language.
- n8n Cloud: thorough, hands-on.
- Agentic AI on LLM APIs (Claude/OpenAI): build and operate.
- Hostinger portals and VPS hosting: independent build and deploy.
- Google Apps Script: robust.
- Java: working understanding; not a primary build language.
- Works directly with non-technical client teams (finance, HR, operations) and explains technical choices in plain language without dropping the technical term.
- Comfortable owning client data under strict confidentiality; disciplined about credentials, access and backups.
- Self-directed on new tools: the stack will keep evolving, and the expectation is to learn, prototype and ship.

Good to have Supabase / PostgreSQL; GitHub and Actions-based CI/CD; Docker and Nginx administration; Streamlit and Replit; REST APIs and webhook security; Power BI; exposure to accounting/compliance data (GST, TDS, inventory).

📌 Automation Engineer (Mumbai)
🏢 Aparna Rajeev Joshi
📍 Mumbai