Upcoming Event: The AI Agent Implementation Framework
Upcoming Event: The AI Agent Implementation Framework
JMARK Business Solutions
Springfield, MO
Managed IT Services (MSP)
EventBridge · AWS Lambda · AWS Step Functions · Amazon SNS · Amazon CloudWatch · AWS Transfer Family · AWS IAM Identity Center · AWS Secrets Manager · Amazon DynamoDB · Amazon Athena · Amazon ECS Fargate
JMARK is a leading managed IT services provider serving businesses across more than 40 states from its headquarters in Springfield, Missouri. When the company’s core data pipeline broke down in December 2025, leaving operational databases disconnected from analytics systems for nearly three months, JMARK needed an urgent, production-safe solution. Avahi designed and delivered a comprehensive AWS-native data analytics platform in just four weeks, restoring data flow across multiple source systems and integrating them into a unified Amazon S3 raw zone. The result was a production-grade, automated pipeline that transformed JMARK’s data freshness from months-stale to near real-time.
Founded in 1988 and headquartered in Springfield, Missouri, JMARK is one of the top managed IT service providers in the United States, ranked in the top 10% of MSPs nationwide. With a team of 115+ professionals and clients in more than 40 states, JMARK delivers fully managed IT services including 24/7 support, cybersecurity, network infrastructure, compliance, and disaster recovery. The company runs a sophisticated internal technology stack, managing a large multi-source data estate spanning SQL Server and MySQL databases, API-connected tools like IT Glue and Auvik for documentation and network monitoring, and manual file-based data workflows. As JMARK scales, turning that data estate into actionable business intelligence has become central to its operational strategy.
In December 2025, JMARK’s Change Data Capture (CDC) pipeline stopped functioning, severing the flow of data from operational databases to its analytics infrastructure. For nearly three months, business intelligence and reporting systems ran on stale data, impairing decision-making across the organization. The affected data estate was substantial: a Microsoft SQL Server instance with hundreds of tables, a MySQL database with more than 800 tables, and API integrations with IT Glue and Auvik, each with different data cadences and operational requirements.
The path to resolution was complicated further when JMARK ruled out native SQL Server replication due to concerns about production stability. That decision eliminated the most straightforward CDC approach and required Avahi to design a custom solution from the ground up, one capable of handling high-volume tables, including those with millions of records, without triggering full reloads or disrupting live systems.
The engagement carried a hard March 31, 2026 sign-off deadline, leaving no margin for delays. On day one of the project, a staffing disruption threatened timelines before development had even begun. JMARK needed a partner who could move fast, adapt under pressure, and deliver a solution built to last.
JMARK already operated within an AWS environment and had an existing commitment to the platform through the AWS Migration Acceleration Program (MAP). AWS offered the breadth of managed data services, from Database Migration Service and Glue to EventBridge, Step Functions, and Transfer Family, that JMARK needed to build a reliable, scalable pipeline without adding operational overhead to their team. The serverless and managed nature of these services directly aligned with JMARK’s goal of minimizing production risk while maximizing automation and observability.
AWS’s native service integration also allowed Avahi to architect a modular pipeline where data sources, refresh schedules, and failure recovery logic could be adjusted without code changes, giving JMARK’s technical team the flexibility to manage and evolve the system long after the engagement closed.
As an AWS consulting partner with deep expertise in data engineering and cloud-native architecture, Avahi was well-positioned to take on a project with tight constraints and zero tolerance for production risk. JMARK needed a partner who understood not just which AWS services to use, but how to sequence them intelligently under real operational pressure, and who could make sound architectural decisions quickly.
Avahi demonstrated exactly that throughout the engagement. The team navigated four major architectural pivots, from Redshift to S3, from Serverless DMS to DMS instances, from Lambda to Glue, and from native CDC to a scheduled reload strategy, without slipping the project timeline. Each pivot was driven by constraints discovered in implementation, and each required a rapid redesign that held. For JMARK, that capacity to adapt without losing momentum was precisely what the engagement required.
Avahi delivered an end-to-end AWS-native data analytics pipeline in four weeks, using Amazon S3 as a hierarchical raw zone organized by data source, schema, and entity, with automated lifecycle policies for cost-optimized long-term storage. Data from SQL Server and MySQL flowed in via AWS Database Migration Service (DMS), while IT Glue and Auvik integrations were handled by AWS Glue jobs. Manual file uploads from JMARK staff connected through AWS Transfer Family, secured with IAM Identity Center integrated with JMARK’s existing Active Directory via SSO, with no separate credential management required.
Because JMARK rejected native SQL Server replication, Avahi designed a hybrid scheduled reload strategy. Smaller tables used Amazon EventBridge cron rules to trigger drop-and-reload jobs via Lambda and DMS, while dozens of large tables, including those with millions of records, used incremental append logic based on timestamp columns to avoid costly full reloads. Schedules and enable/disable controls were fully configurable from the AWS console without touching code, giving JMARK’s team operational flexibility from day one.
For MySQL, DMS handled full load plus ongoing replication, achieving near real-time data synchronization. A checker Lambda function polled replication task status continuously and automatically restarted failed tasks, a capability proven during the engagement when a MySQL server outage caused a replication interruption and the pipeline self-recovered without manual intervention.
API integrations for IT Glue and Auvik were built as production-grade AWS Glue jobs with pagination handling, rate limiting, and AWS Secrets Manager authentication. The Auvik integration ran as an AWS Step Functions orchestration with retry logic and Amazon SNS failure alerts on automated refresh schedules. IT Glue data ingested daily via Amazon ECS Fargate tasks, with Amazon DynamoDB tracking processing metadata for pipeline observability and future alerting extensibility.
Data validation was performed using AWS Glue Python shell jobs, which scaled to profile the full data estate of 2,416 tables after Lambda functions timed out on the same workload. Reconciliation statistics comparing source row counts to S3 counts were generated and shared with JMARK, confirming data integrity across all migrated sources. All infrastructure was tagged for MAP program compliance, code was handed off to JMARK’s GitHub repository via pull requests, and multiple knowledge transfer sessions were conducted with JMARK’s technical lead to enable independent operation and maintenance after the engagement closed.
Avahi delivered a complete, production-grade data analytics pipeline in four weeks, meeting the March 31 deadline with time to spare, despite a day-one staffing disruption and four major architectural pivots mid-delivery. JMARK’s data infrastructure went from a months-long standstill to a continuously refreshing, multi-source analytics foundation with automated failure recovery and end-to-end observability.
With all data paths validated and zero events in the dead-letter queue at project completion, JMARK’s analytics team gained reliable access to current data across every integrated source, laying the groundwork for data-driven decision-making at scale.
Let’s explore your high-impact AI opportunities together in a complimentary session
