AI Systems Engineer

Iyyappan Mani

Building practical AI systems for operations, automation, and agent-native workflows.

Portfolio of internal tools, MCP servers, support automation systems, and AI-assisted product experiments grounded in SAP, ERP, data workflows, and real operational constraints.

Open to roles Dublin, Ireland
AI operations systems SAP / ERP automation RAG + knowledge systems MCP servers Agent infrastructure Workflow apps Product prototypes

Selected work

Projects built in the transition, not around it.

01 PRODUCTION STAGING
NORA

AI-assisted support operations system for Outlook-to-ERP logging, AWS-routed email understanding, My Calls, Team Chat, Knowledge Library, admin settings, and controlled execution.

Outlook ERP API Postgres LLM Advisory Portal UI
02 PRIVATE DEMO READY
BART

Hosted marketing lead intake automation for Enterpryze workflows, turning the clean new-prospect handoff from a 15-20 minute manual delay into an around-five-second Teams alert.

Enterpryze API Jotform Teams Bot SQLite TypeScript
03 LIVE PRIVATE PILOT
POC App

Warehouse proof-of-collection app for SAP Business One supplier returns, with digital courier signatures, signed PDF generation, and SAP UDF writeback.

SAP B1 Service Layer SQL Server TypeScript Fastify PDFKit
04 FINAL BUILD STAGE
ARGUS

RAG-based support assistant for business systems knowledge, built to learn retrieval pipelines and grounded AI responses.

Python FastAPI Postgres pgvector RAG
05 PLANNING / LOCAL PILOT
GRODD

Installable Peppol submission product for SAP Business One invoices and credit notes, built around a canonical document model.

SAP B1 SQL Peppol Canonical Model Local Dashboard SQLite
06 PUBLIC DEMO READY
Enterpryze MCP

Hosted MCP server for Enterpryze sales-order workflows, with read tools, analysis tools, guarded create flow, public setup docs, and bearer-protected access.

MCP TypeScript Fastify Enterpryze API Cloudflare Tunnel
07 PHASE 4 PLANNED
THAWNE

Autonomous ecosystem intelligence agent tracking developments, services, and signals across the Tempo ecosystem.

MCP REST Automation Knowledge Graph Tempo
08 LIVE
DIBNY

Collaborative art game where AI agents pay to draw on a shared canvas.

MCP MPP x402 Tempo

About

Grounded in systems. Built in collaboration.

My background is in SAP Business One, business systems, reporting, and data workflows, where I spent years turning operational needs into working solutions for users and stakeholders.

That work taught me how to think in systems, not features. I learned how processes break, where teams get blocked, and what useful software actually has to do once it leaves the whiteboard.

More recently, I have been building AI-assisted products and experimenting with agent-driven systems, with a focus on making AI genuinely useful in real-world environments.

IM x AI reflects how I work now: human judgment, direction, and taste, strengthened by serious collaboration with AI during research, prototyping, implementation, and iteration.

I am most interested in work that sits close to infrastructure, users, and operational reality: the layer where product thinking, technical depth, and practical automation meet.

Experience

Milner Browne

SAP Business One Consultant

Delivering SAP Business One implementations, reporting systems, and data workflow solutions across ERP, MongoDB, and analytics environments

2018 - Now

Groupon

Customer Support Representative

Built a strong foundation in troubleshooting, communication, and customer-facing problem solving in a high volume support environment

2014 - 2016

Hackathons

Notion MCP Hackathon

Participated in the Notion MCP Challenge by shaping and testing THAWNE's bidirectional Notion interface, enabling human review and structural intelligence flow for an autonomous blockchain agent

2026

Certification

SAP Certified Application Associate

2025

Open to the right conversation.

If you are hiring for AI systems, agent tooling, or technical product roles, I would be glad to connect.