Skip to content

Commit 6a692c3

Browse files
committed
feat(agentic idp): First version of agentic IDP using strands
1 parent ab7e48f commit 6a692c3

File tree

22 files changed

+10531
-1720
lines changed

22 files changed

+10531
-1720
lines changed

config_library/pattern-2/lending-package-sample/config.yaml

Lines changed: 100 additions & 97 deletions
Large diffs are not rendered by default.

docs/extraction.md

Lines changed: 261 additions & 84 deletions
Large diffs are not rendered by default.

docs/govcloud-deployment.md

Lines changed: 29 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -10,6 +10,7 @@ The GenAI IDP Accelerator now supports deployment to AWS GovCloud regions throug
1010
## Architecture Differences
1111

1212
### Standard AWS Deployment
13+
1314
```mermaid
1415
graph TB
1516
A[Users] --> B[CloudFront Distribution]
@@ -22,6 +23,7 @@ graph TB
2223
```
2324

2425
### GovCloud Deployment
26+
2527
```mermaid
2628
graph TB
2729
A[Direct S3 Upload] --> F[Core Processing Engine]
@@ -75,76 +77,85 @@ aws cloudformation deploy \
7577
The following services are automatically removed from the GovCloud template:
7678

7779
### Web UI Components (22 resources removed)
80+
7881
- CloudFront distribution and origin access identity
7982
- WebUI S3 bucket and build pipeline
8083
- CodeBuild project for UI deployment
8184
- Security headers policy
8285

8386
### API Layer (20+ resources removed)
87+
8488
- AppSync GraphQL API and schema
8589
- All GraphQL resolvers and data sources
8690
- 10+ Lambda resolver functions
8791
- API authentication and authorization
8892

8993
### Authentication (8 resources removed)
94+
9095
- Cognito User Pool and Identity Pool
9196
- User pool client and domain
9297
- Admin user and group management
9398
- Email verification functions
9499

95100
### WAF Security (6 resources removed)
101+
96102
- WAF WebACL and IP sets
97103
- IP set updater functions
98104
- CloudFront protection rules
99105

100106
### Analytics Features (8 resources removed)
107+
101108
- Analytics processing functions
102109
- Knowledge base query functions
103110
- Chat with document features
104111
- Text-to-SQL query capabilities
105112

106113
### HITL Support
114+
107115
- SageMaker A2I Human-in-the-Loop
108116
- Private workforce configuration
109117
- Human review workflows
110118

111-
112119
## Core Services Retained
113120

114121
The following essential services remain available:
115122

116123
### Document Processing
124+
117125
- ✅ All 3 processing patterns (BDA, Textract+Bedrock, Textract+SageMaker+Bedrock)
118126
- ✅ Complete 6-step pipeline (OCR, Classification, Extraction, Assessment, Summarization, Evaluation)
119127
- ✅ Step Functions workflows
120128
- ✅ Lambda function processing
121129
- ✅ Custom prompt Lambda integration
122130

123131
### Storage & Data
132+
124133
- ✅ S3 buckets (Input, Output, Working, Configuration, Logging)
125134
- ✅ DynamoDB tables (Tracking, Configuration, Concurrency)
126135
- ✅ Data encryption with customer-managed KMS keys
127136
- ✅ Lifecycle policies and data retention
128137

129138
### Monitoring & Operations
139+
130140
- ✅ CloudWatch dashboards and metrics
131141
- ✅ CloudWatch alarms and SNS notifications
132142
- ✅ Lambda function logging and tracing
133143
- ✅ Step Functions execution logging
134144

135145
### Integration
146+
136147
- ✅ SQS queues for document processing
137148
- ✅ EventBridge rules for workflow orchestration
138149
- ✅ Post-processing Lambda hooks
139150
- ✅ Evaluation and reporting systems
140151

141-
142152
## Access Methods
143153

144154
Without the web UI, you can interact with the system through:
145155

146156
### 1. Direct S3 Upload
147-
```bash
157+
158+
````bash
148159
# Upload documents directly to input bucket
149160
aws s3 cp my-document.pdf s3://InputBucket/my-document.pdf
150161

@@ -154,26 +165,30 @@ Using the lookup script
154165
```bash
155166
# Use the lookup script to check document status
156167
./scripts/lookup_file_status.sh documents/my-document.pdf MyStack
157-
```
168+
````
158169
159170
Or navigate to the AWS Step Functions workflow using the link in the stack Outputs tab in CloudFormation, to visually monitor workflow progress.
160171
161-
162172
## Monitoring & Troubleshooting
163173
164174
### CloudWatch Dashboards
175+
165176
Access monitoring through CloudWatch console:
177+
166178
- Navigate to CloudWatch → Dashboards
167179
- Find dashboard: `{StackName}-{Region}`
168180
- View processing metrics, error rates, and performance
169181
170182
### CloudWatch Logs
183+
171184
Monitor processing through log groups:
185+
172186
- `/aws/lambda/{StackName}-*` - Lambda function logs
173187
- `/aws/vendedlogs/states/{StackName}/workflow` - Step Functions logs
174188
- `/{StackName}/lambda/*` - Pattern-specific logs
175189
176190
### Alarms and Notifications
191+
177192
- SNS topic receives alerts for errors and performance issues
178193
- Configure email subscriptions to the AlertsTopic
179194
@@ -182,6 +197,7 @@ Monitor processing through log groups:
182197
The following features are not available:
183198
184199
### ❌ Removed Features
200+
185201
- Web-based user interface
186202
- Real-time document status updates via websockets
187203
- Interactive configuration management
@@ -192,6 +208,7 @@ The following features are not available:
192208
- Document knowledge base chat interface
193209
194210
### ✅ Available Workarounds
211+
195212
- Use S3 direct upload instead of web UI
196213
- Monitor through CloudWatch instead of real-time UI
197214
- Edit configuration files in S3 directly
@@ -204,18 +221,21 @@ The following features are not available:
204221
## Best Practices
205222
206223
### Security
224+
207225
1. **IAM Roles**: Use least-privilege IAM roles
208226
2. **Encryption**: Enable encryption at rest and in transit
209227
3. **Network**: Deploy in private subnets if required
210228
4. **Access Control**: Implement custom authentication as needed
211229
212230
### Operations
231+
213232
1. **Monitoring**: Set up CloudWatch alarms for critical metrics
214233
2. **Logging**: Configure appropriate log retention policies
215234
3. **Backup**: Implement backup strategies for important data
216235
4. **Updates**: Plan for template updates and maintenance
217236
218237
### Performance
238+
219239
1. **Concurrency**: Adjust `MaxConcurrentWorkflows` based on load
220240
2. **Timeouts**: Configure appropriate timeout values
221241
3. **Memory**: Optimize Lambda memory settings
@@ -226,11 +246,13 @@ The following features are not available:
226246
### Common Issues
227247
228248
**Missing Dependencies**
249+
229250
- Ensure all Bedrock models are enabled in the region
230251
- Verify IAM permissions for service roles
231252
- Check S3 bucket policies and access
232253
233254
**Processing Failures**
255+
234256
- Check CloudWatch logs for detailed error messages
235257
- Verify document formats are supported
236258
- Confirm configuration settings are valid
@@ -254,13 +276,14 @@ If migrating an existing deployment:
254276
## Cost Considerations
255277
256278
GovCloud pricing may differ from commercial regions:
279+
257280
- Review [GovCloud Pricing](https://aws.amazon.com/govcloud-us/pricing/)
258281
- Update cost estimates in configuration files
259282
- Monitor actual usage through billing dashboards
260283
261284
## Compliance Notes
262285
263286
- The GovCloud version maintains the same security features
264-
- Data encryption and retention policies are preserved
287+
- Data encryption and retention policies are preserved
265288
- All processing remains within GovCloud boundaries
266289
- No data egress to commercial AWS regions

0 commit comments

Comments
 (0)