iTromsø · IBM, VC2
10 minutes per day — time spent reviewing potential stories
Independently reportedMedia & entertainment · Document processing · NLP · In production
Document processing
Affinda · AWS
Vendor-reportedTechnology & softwareIn productionLLMLive 2024
Vendor-reported. The customer is named and the numbers are quoted from the source page, but the account comes from the vendor. No independent confirmation.
Reported by amazon.com (vendor self-report). Checked against the source page on 2026-09-06. 2 of 2 figures below appear on that page word for word.
The problem
Traditional machine learning models for document data extraction required extensive technical setup and data annotation, making the process time-consuming and complex for both Affinda and its customers.
What was deployed
- Vendor
- AWS
- Products
- Amazon Bedrock, Amazon SageMaker, Amazon Elastic Kubernetes Service (Amazon EKS), Amazon Elastic Compute Cloud (Amazon EC2), AWS CloudFormation
- Models
- Claude Sonnet 3.5 v2, Claude 3.7 Sonnet, Claude Sonnet 4
- Technique
- LLM
- Build or buy
- Bought off the shelf
- Deployment
- SaaS
- Data used
- small number of idiomatic examples that show the LLM in its context window what the problem is; user input
- Scale
- organizations worldwide; across industries and use cases; straight-through processing at scale
What changed
Each row is quoted from the source. Figures we could not find on the page in those words are marked — they are kept, not deleted, so you can judge them.
setup time for new use cases
“cutting setup time for new cases by 90 percent”
Quoted word for word from the source · round percentage, no baseline given
cost for product delivery team to roll out a custom use case
“reducing costs by 90 percent”
Quoted word for word from the source · round percentage, no baseline given
improvedcustomer experience
“improving customer experience”
Quoted word for word from the source
cutengineering overhead for model correction
“simplified model correction through natural-language explanations, cutting engineering overhead.”
Quoted word for word from the source
minutestime for customers to configure new data extraction modelsfrom weeks or even months
“The platform now allows customers to self-serve, configuring new data extraction models themselves in a matter of minutes—a task that previously took Affinda’s internal team weeks or even months.”
Quoted word for word from the source
reducedtime to value for Affinda’s customers
“This approach has reduced time to value and increased return on investment for Affinda’s customers.”
Quoted word for word from the source
increasedreturn on investment for Affinda’s customers
“This approach has reduced time to value and increased return on investment for Affinda’s customers.”
Quoted word for word from the source
improvedaccuracy
“customers also benefit from instant learning and adaptation, which improves accuracy and allows for straight-through processing at scale.”
Quoted word for word from the source
ROI as stated: increased return on investment for Affinda’s customers.
“AWS is our preferred cloud provider, and technologies like Amazon SageMaker and Amazon Bedrock give us confidence in data sovereignty, regional data processing, and seamless integration with our existing AWS environment.”
Difficulties and limits
- Initially relied on traditional machine learning (ML) models that required extensive technical setup and data annotation, making the process time-consuming and complex for both Affinda and its customers.
- Used open-source models that were fine-tuned for specific document types and schemas, which took a lot of engineering effort.
- Even with a self-service model, customers still needed to produce their own training data sets.
Sources
- amazon.com — vendor self-report
Similar deployments
Amplitude · Wolfia
4X — throughput of security questionnaires
Vendor-reportedTechnology & software · Document processing · AI agents · Scaled
Adyen · Rossum
70 — automation rate
Vendor-reportedBanking & finance · Document processing · NLP · Scaled · 2023
Skaylink · Moterra
dropped by around 80% — NIS2 preparation time
Vendor-reportedTechnology & software · Document processing · LLM · Scaled
OM Pharma · DrugCard
3 minutes — review time per article
Vendor-reportedPharma & biotech · Document processing · NLP · In production
AMPREP (Amalgamations Repco) · Cimplify.AI
100% — invoice automation
Vendor-reportedManufacturing · Document processing · NLP · In production