customer_support · public · workflow
Boti: Agentic AI assistant on Amazon Bedrock transforms citizen access to Buenos Aires government procedures
Citizens of Buenos Aires faced difficulty navigating the city's complex bureaucratic landscape of over 1,300 government procedures, each with its own logic, nuances, and exceptions, making it hard to find the right procedure quickly.
How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · Citizen submits question
The process begins when a user submits a question, which is passed in parallel to both the input guardrail system and the government procedures agent.
Tools used
Amazon Bedrock Converse APIAmazon Titan Text Embeddings v2Cohere Multilingual v3Claude 3.5 SonnetClaude 3 SonnetHaiku 3
Outcome
The agentic system achieves up to 98.9% top-1 retrieval accuracy, a 12.5–17.5% improvement over standard RAG methods, blocks 100% of harmful queries, and Boti facilitates more than 3 million conversations each month.
What failed first
Standard retrieval-augmented generation approaches struggled to disambiguate similar government procedures, and the mixture of chunks they returned increased the likelihood of generating incorrect responses.
Results
Time savedmore than 3 million conversations each month
Volumeover 1,300
Running sinceFebruary 2019
Grounding & classification
Source type: technical build writeup
38 fields verified against source quotes, 4 dropped as unverifiable.
agentic workflowai agentcontent generationconversational airagsentiment analysisknowledge basepolicy documenthuman review describedmetric backednamed customerproduction runtime claimedtools describedworkflow describedgovernmentaccuracy improvementdeflection ratethroughput increasetechnical build writeupback office opscustomer supportintake to triagerag answering