Faris Hussain

Case Studies

How I approach data platform and agentic AI problems

Two projects, in more depth than the summary cards — the problem I was solving, the approach I took, and the outcome it produced.

Enmacc GmbH · Feb 2025 – Present

Energy Trading Agents — Development & Research

Google ADK · Amazon Bedrock · AWS Glue/S3/Lambda · Airflow · Terraform

The challenge

Energy-trading agents need domain-specific workflows and a way to assess negotiation behaviour. The work includes agent-development R&D at Enmacc and a separate academic comparison of LLM-driven and rule-based agents in simulated request-for-quote (RFQ) negotiations.

The approach

  • At Enmacc, designed goal-oriented energy-trading agents using Google ADK and Amazon Bedrock as part of agent-development R&D.
  • Platform contributions included AWS Glue, S3, Lake Formation and Lambda data foundations, with Terraform and CloudFormation for infrastructure as code.
  • Implemented Apache Airflow for workflow orchestration and visibility into pipeline health.
  • Separately, the M.Eng. research compared LLM-driven and rule-based agents in energy RFQ simulations, with energy-specific agents, data transformations, negotiation workflows and experiment analytics.

The outcome

The contributions cover agent-development R&D and platform engineering at Enmacc, alongside a separate simulation-based research evaluation. The academic work examines negotiation outcomes, economic measures and model telemetry; it does not establish live autonomous trading performance, commercial returns or adoption of the findings in production.

Royal Cyber Inc. · Mar 2023 – Nov 2024

Data Vault 2.0 Modernisation for Insurance & Finance

Snowflake · Data Vault 2.0 · VaultSpeed · Talend · Kafka · Airflow · SAP

The challenge

Insurance and finance data held in Mainframe/DB2 and SAP systems needed to be integrated into modern enterprise warehouses with consistent modelling across multiple sources.

The approach

  • Migrated Mainframe/DB2 insurance data to Snowflake and designed Data Vault 2.0 models for SAP finance data.
  • Built SQL transformations and pipelines using Talend for batch processing and Kafka for streaming.
  • Automated orchestration with Airflow and warehouse modelling with VaultSpeed, integrating multiple source systems.
  • Replaced SAP BODI ETL logic with Talend and managed warehouse and pipeline deployments through Azure DevOps CI/CD.

The outcome

Integrated legacy insurance and finance data into enterprise warehouse models and replaced SAP BODI ETL logic with Talend. The delivered work covered source migration, modelling, pipeline orchestration and deployment automation.

Have a similar problem?

Let's talk about your data platform or agentic AI challenge.

Book a call