Client Overview
Datamaran is a global leader in ESG (Environmental, Social, and Governance) data analytics, providing AI-driven insights for corporate risk management and sustainability reporting. The platform processes thousands of reports, regulations, and news items daily, leveraging advanced NLP models like FinBERT-FLS for ESG sentiment classification, AWS Cloud infrastructure, and secure data pipelines.
With its growing data and AI workloads, Datamaran needed to ensure resilience, scalability, and cost efficiency across its AWS-based infrastructure while maintaining compliance with internal data protection and disaster recovery standards.
Challenge
As the company expanded its AI pipelines and Bedrock-based NLP workloads, several critical needs emerged:
- Ensuring Business Continuity across multiple AWS regions and services
- Building Disaster Recovery (DR) and Backup Automation in alignment with internal policies
- Optimizing Cloud Costs, particularly for PostgreSQL, S3 backups, SageMaker, and EC2 usage
- Standardizing Infrastructure as Code (IaC) for faster and more secure deployment
- Enhancing Observability and Incident Readiness, reducing potential downtime.
Solutions Delivered by Gart Solutions
Disaster Recovery Architecture Implementation
Gart Solutions designed and implemented a multi-regional DR setup using Terraform and AWS services in accordance with Datamaran’s internal Disaster Recovery Plan
Key deliverables included:
- Replication strategy for S3 and RDS PostgreSQL across primary and DR sites
- Automated environment rebuilds scripts tested to restore production infrastructure in under 2 hours for database failure and 5 minutes for client-facing app failover.
- CloudFront and WAF configuration for secure content delivery and global traffic management
Infrastructure as Code & CI/CD Enhancements
- Developed Terraform templates for:
- AWS CloudFront distribution and security groups.
- ECS modules and dynamic API infrastructure (DMX API repository).
- Lambda configurations, parameterized for scalability and easier updates.
- Integrated pipelines in Git-based CI/CD (Gart-managed repository) to deploy and test dynamic infrastructure in isolated environments
Cloud Cost Optimization
Through detailed cost breakdown and service-level review:
- Identified high-cost contributors including RDS PostgreSQL ($1,721/month) and SageMaker ($1,452/month) workloads.
- Implemented Savings Plans and Spot Instances to optimize EC2 utilization, achieving an estimated 20–25% monthly cost reduction
- Introduced cost-tracking dashboards for ongoing visibility and forecasting.
Backup Policy Modernization
- Enhanced backup automation for MongoDB Atlas, ElasticSearch, and Cognito users in AWS.
- Ensured compliance with internal data retention policies — hourly, daily, and weekly snapshots encrypted in transit and at rest
Incident Response and Monitoring Integration
- Integrated AWS CloudWatch and Inspector for proactive vulnerability detection and logging.
- Established incident reporting workflow tied to Datamaran’s InfoSec policy and Slack/Email escalation channels
- Improved audit readiness with centralized reporting via AWS Security Hub.
Key Technologies
- AWS Services: RDS (PostgreSQL), S3, EC2, ECR, ECS, CloudFront, CloudWatch, SageMaker, Inspector, Security Hub, WAF, Bedrock.
- DevOps Tools: Terraform, GitLab CI/CD, Docker, Argilla, Vellum, spaCy, Anthropic Claude.
- Infrastructure Scope: Multi-region (EU/Ireland, US East), VPC-based architecture with automated DR failover.
