Saiyudh MannanLet's talk

AI Automation Engineer · Saiyudh Mannan

AI automations that run your business, safely.

I design and build the workflows, agents and data pipelines that take repetitive work off your team, with the access controls and monitoring that let you trust them.

Let's talkSee the work
The CitadelApproach

How an automation runs.

  1. 01

    Trigger

    A form, an email, a new row or a schedule starts the run.

  2. 02

    AI step

    A model reads, classifies or drafts, and its output is checked before anything acts on it.

  3. 03

    Action

    Your CRM, inbox, sheet or database is updated through scoped credentials.

  4. 04

    Monitor

    Every run is logged, and failures alert a person instead of failing silently.

Secure by design.

  • Least privilege. Each automation gets only the access it needs, nothing shared or admin-wide.
  • Secrets stay secret. Keys live in a vault, never in code or prompts, and untrusted input is checked before an AI step can act on it.
  • Nothing runs unseen. Logs, alerts and a hand-over document, so your team can audit or change it later.
The ArchiveWork

Systems, not demos.

Case study 01

Sidekick — AI Personal Co-Worker

Problem
Owners lose hours every week to research, bookings and reminders that follow the same steps each time.
Built
An AI co-worker built with LangChain and LangGraph that searches the web, books travel through Amadeus and sends WhatsApp updates through Twilio.
Result
Routine admin is requested in plain English, and each task is checked against its success criteria before the result is reported back.

LangChain · LangGraph · Gradio · Playwright · Twilio · Python

View the code ↗

Case study 02

DevOps Automation — Terraform & AWS

Problem
Growing teams need reliable infrastructure, but manual deployments are slow and every release risks downtime.
Built
A production AWS environment defined in Terraform: Docker services on ECS behind a load balancer, with RDS, shared storage and CloudWatch monitoring.
Result
Every code change flows through a CI/CD pipeline and deploys with zero downtime, the same way every time.

Terraform · ECS · RDS · EFS · Docker · GitHub Actions · GitLab CI

View the code ↗

Case study 03

Real-Time Patient Flow Analytics — Azure

Problem
Operational reports built by hand are already out of date by the time anyone reads them.
Built
A real-time pipeline on Azure: Event Hub streams events into Databricks, where PySpark refines them through bronze, silver and gold layers into Power BI.
Result
Patient-flow KPIs update live on a dashboard, so bottlenecks show up during the shift instead of in next week's report.

Azure Event Hub · Databricks · PySpark · Azure Synapse · Power BI

View the code ↗

More projects

The ForgeAbout

Engineer first, automation second.

Before AI automation I spent years as a data engineer, building ETL, CI/CD and monitoring for production systems. That is why the automations I build are readable, tested and observable, not just impressive in a demo.

I work remotely with teams anywhere, from a first workflow to a system your staff rely on every day.

4+
years building data systems
Remote
based in Germany
EN
working language
  • n8n
  • AI agents
  • LLM workflows
  • Data pipelines
  • AWS · Azure
  • Terraform

Experience

  1. Jan 2025-CurrentAI Automation Engineer -- FreelancerSaaS Applications with n8n, RAG Chatbots with Langchain and Supabase on Node.js and FastAPI.
  2. Mar 2023–Sep 2024Data Engineer — Adastra GmbH, MagdeburgTeradata + AWS ETL automation · CI/CD & monitoring as code (Terraform, CloudWatch, Control-M)
  3. May 2022–Feb 2023Part-Time Data Engineer — Adastra GmbHCI/CD on IONOS Kubernetes + AWS · Python automation & monitoring
  4. Jun 2018–Dec 2018Oracle DBA Engineer — Dhaka Bank, DhakaAutomated SQL financial reports in Python

Education

  1. 2026-CurrentM.Sc. Digital EngineeringOtto von Guericke University, Magdeburg
  2. 2026Data Science & AI TrainingLe Wagon
  3. 2024–2025SAP — ABAP · BW/S4HANATH Brandenburg · ERP4STUDENTS
  4. 2024–2025German A1–B2 (B1 certificate)VHS Magdeburg
  5. 2019–2025M.Sc. Data & Knowledge EngineeringOtto von Guericke University, Magdeburg
  6. 2013–2018B.Sc. Computer Science & EngineeringNorth South University, Dhaka

Certifications

  • SAPSAP BW/4HANA — Data Modeling & LumiraTH Brandenburg · 2025
  • SAPSAP BW/4HANA — Dataflow & ReportingTH Brandenburg · 2025
  • SAPIntroduction to ABAPTH Brandenburg · 2025
  • DATAGoogle Data Analytics ProfessionalGoogle / Coursera · 2021
  • CLOUDAWS Cloud Practitioner (CLF-02)Amazon Web Services · 2024
  • CLOUDDocker, Kubernetes & AWS BootcampUdemy · 2025
  • AILLM Engineer BootcampEd Donner Academy · 2025
The VaultContact

Let's talk.

Tell me which task eats up your team's week, and we'll work out whether it can be automated.

Magdeburg, Germany