CloudThat

cloudthat.com

CloudThat combines advisory work, hands-on technology implementation and ongoing operations. Buyers can commission architecture reviews and roadmaps separately from application changes, AI integration or infrastructure management. Its AI practice explicitly covers customized machine-learning and natural-language-processing solutions, rather than relying on training or product marketing to establish delivery capability. Corporate education and the packaged GenAI STACQ solution should be evaluated separately from consulting scope and production-system ownership. ^[1] ^[2] ^[3] ^[4] ^[5] ^[6]

Services and delivery

AI strategy and opportunity assessment

Advisory. Uses interviews, workshops and data exploration to identify AI opportunities and develop a roadmap before model development or integration. ^[2]

AI/ML development and integration

Implementation. Builds customized machine-learning and natural-language-processing solutions, integrates them into existing systems, and provides post-implementation optimization and technical support. ^[2]

Generative-AI model customization

Implementation. Covers requirements gathering, model selection, development, fine-tuning, testing and validation. Advertised applications include contextual chat, document search, assessment generation and audio/video summarization, with an emphasis on educational uses and AWS deployment. ^[7]

Cloud migration assessment and planning

Advisory. Assesses infrastructure, applications and data; develops migration roadmaps and modernization strategies; and provides architectural optimization and guidance through implementation. ^[8]

Backup and disaster-recovery implementation

Implementation. Implements backup and disaster-recovery arrangements within its cloud migration offering. Buyers should define recovery requirements and testing responsibilities in the project scope. ^[8]

Application modernization

Implementation. Rehosts, refactors, re-platforms and rearchitects legacy applications. Delivery includes source-code changes, database migration, deployment, testing and documentation. ^[3]

Data engineering and analytics implementation

Implementation. Integrates data sources, cleans and prepares data, migrates data, manages databases and pipelines, and develops visualizations. This offering establishes data-platform delivery, not AI/ML delivery by itself. ^[9]

Managed cloud infrastructure and DevOps

Managed services. Provides ongoing infrastructure management with 24/7 monitoring, automated provisioning, configuration, patching, scaling and performance tuning. The scope also includes CI/CD, infrastructure as code, backup, recovery and data-lifecycle management. ^[4]

Managed cloud security operations

Managed services. Offers security monitoring, incident detection, response and recovery through a Cloud Security Operations Center. ^[4]

AWS Well-Architected review

Advisory. Reviews AWS workloads and produces a findings report with tailored best practices and remediation priorities. The review does not establish that remediation implementation is included. ^[1]

Corporate technology training

Non-technical service. Provides customized corporate courses, assessments, hands-on labs and capstone projects. This is education rather than ownership of production-system implementation. ^[5]

GenAI STACQ packaged solution

Software product. An Amazon Bedrock-based solution for summarization, topic extraction, assessment generation and contextual question-answering. Its page describes serverless, pay-as-you-go architecture without clearly separating licensing, cloud consumption and professional-services charges. ^[6]

Company details

Industries and client fit

Published delivery examples span asset management, workspace discovery, health technology and vehicle-data analytics. These establish relevant project experience, rather than comprehensive specialization across those industries. The generative-AI service page also describes educational applications. ^[10] ^[11] ^[12] ^[13] ^[7]

Track record

For Kotak Mahindra Asset Management Company, CloudThat describes implementing generative-AI email responses using Claude v2, SageMaker preprocessing, DynamoDB, Power Automate, API Gateway and Lambda. It reports faster query resolution, but the supplied evidence contains no quantified improvement or evaluation methodology. ^[10]

For Qdesq, CloudThat reports a multi-agent platform for workspace recommendations, lead capture and visit scheduling. The architecture used Amazon Bedrock AgentCore, Strands Agents, Claude 3.5 Sonnet, FastAPI on EC2, DynamoDB and CRM integration. Its reported 98% recommendation accuracy lacks an explained test set or measurement procedure in the supplied evidence. ^[11]

The MedPay case describes moving microservices from Elastic Beanstalk to Amazon EKS, provisioning networks and clusters, and implementing Argo CD delivery, autoscaling, Prometheus/CloudWatch monitoring, OpenSearch logging and Grype image checks. CloudThat reports improved availability and scalability. ^[12]

For Ultraviolette, the reported data platform ingested vehicle data through AWS IoT Core and MSK, processed it with Lambda and Logstash, and used S3, OpenSearch, Glue and Redshift. Amazon QuickSight supplied dashboards. This is evidence of streaming-data and analytics implementation, not independently of machine learning. ^[13]

Scope and limitations

The supplied evidence establishes commercial AI/ML delivery, but does not establish standard prices, project durations or independently validated outcome guarantees.
GenAI STACQ licensing, cloud consumption and professional-services charges are not clearly separated in the supplied notes.
The dedicated DevOps/DevSecOps page was not consistently retrievable. DevOps scope here relies on the managed-services page and MedPay delivery case.
General AI language in cloud migration and managed services does not establish a separately scoped AI service within those offerings.

Questions to ask

Which deliverables are advisory, implementation, training or ongoing operations, and who owns each workstream?
What source-code rights, documentation, deployment assets and handover support are included?
Which comparable production customers can provide references for the proposed architecture and scope?
For AI systems, what evaluation criteria, human-review controls, retention rules and model-provider constraints apply?
What support coverage, incident responsibilities, recovery targets and service levels will be contractual?
Which remediation tasks, cloud charges and model-consumption costs are excluded from the quoted fees?

Overview