endorsedCompleted

Generative & Agentic AI Platform for Blockchain Company

Services Covered:

AI and Machine Learning • Custom Software Development • API Development

Industry

Government & Public Safety

Duration

30 months

Budget

200K - 999K

Start Date

1 July 2022

Team Size

6-10 Employees

Engagement Model

Offshore

Project Capability Score (PCS)

75/100

Good

Project Capability Score (PCS) estimates how capable Codewinglet Private Limited is of successfully delivering a project like this, based on its past experience, track record, and ability to handle similar work. Learn more about PCS

Client

Confidential

Newark, United States of America

Project Summary

Codewinglet Private Limited delivered generative and agentic AI services to an AI-powered blockchain company, building text-generation, summarization, structured-data extraction and RAG pipelines, and developing production Python microservices. They implemented agent orchestration with Temporal.io and LangChain-based frameworks, integrated services with the Azure ecosystem, and provided CI/CD, observability, and developer documentation. The client reported high-quality results, fast delivery, responsiveness, and adaptability.

Key Challenges

Generative AI Services
Agentic AI Services
Fast Python AI Service Development

Project Deliverables

Generative AI microservices (FastAPI APIs)
RAG pipeline with document processing
Structured field evaluation framework
Agentic AI framework with LangChain/LangGraph
EMS triage, recruitment, and workforce agents
State tracking and audit logs
Python AI microservices with DB integration
CI/CD pipelines and observability tools
Developer docs and knowledge transfer

Project Solution

Built generative AI microservices (FastAPI) for text generation, summarization, and structured data extraction; implemented RAG pipelines for document ingestion and context-aware responses; developed agentic AI orchestration using LangChain/LangGraph and Temporal.io with state tracking and audit logs; and rapidly delivered production-grade Python microservices integrated with Azure (Cosmos DB, Service Bus, Redis), along with CI/CD, observability and developer handover.

Project Outcome

Ability to ingest and process 500+ page PDFs with >95% accuracy in field extraction
RAG response relevance >90% in test queries
AI response latency averaged <2 seconds
Agent workflow completion rate >85% without manual intervention
EMS agent triage accuracy >90% against reference cases
Temporal workflows with <1% failed retries
New AI services delivered to staging in <2 days on average
System uptime >99% for deployed AI microservices
Scalability to handle 1,000+ concurrent requests without degradation
Consistent on-time delivery and high responsiveness

Platforms

Cloud

Tech Stack

Fast API
Azure Cosmos DB
Azure
Redis
Python

Client Endorsement

Overall Review Rating

Timeliness

Cost Rating

Willing to Refer

Quality of Deliverables

“The client was pleased with the project's results. Codewinglet Private Limited was efficient — they always delivered items on time or ahead of schedule without compromising quality. Moreover, they were quick to adapt to changes, and their AI expertise and efficient execution impressed the client.”
Anonymous
Project Manger
Date: 08/27/2025