Karthik Jothi

AI Modules

• Spearheaded the analysis and planning of AI features to enhance product usability and overall value.
• Designed and developed microservices from the ground up using .NET Core, Azure Function Apps, and Semantic Kernel frameworks.
• Integrated Azure Open AI for LLM (Large Language Model) support.
• Led a team of 5 engineers in the development and maintenance of AI-driven projects.
• Post-AI integration, ticket resolution time was reduced by 35%, allowing engineers to focus more on critical solutions.

Key Contributions:

Copilot: Developed a conversational assistant to expedite ticket resolution for both customers and agents using an agent workflow capable of multiple task automation. Implemented skills such as answering customer queries using Retrieval-Augmented Generation (RAG).
Provided agents with suggested responses, ticket analytics, and synthetic data such as sentiment and intent.
Automated macro suggestions for increased efficiency.
Content Generator: Created a tool to assist users in generating optimized blog posts and knowledge base articles based on user inputs.
Review Generator: Developed a review system to evaluate agent responses, highlighting addressed and unaddressed points, while offering suggestions and metrics.
Writing Assistance: Built features for content creation, including summarization, expansion, tone modification, and grammar checks.
Natural Language Actions: Enabled users to perform actions like search, filtering, and ticket creation through conversational interactions with the system.

Tenant Manager

• Architected and developed a microservice project from scratch, using .NET Core APIs, for managing tenant onboarding, resource creation, configurations, and social sign-ups.
• Implemented functionalities for tenant maintenance, such as upgrades, status monitoring, opportunity tracking, and billing management.

Email Sender System

• Designed and developed a queue-based email system using Azure Function Apps, with integration to AWS SES and SNS for email delivery and failure handling.
• Added support for SMTP server configurations to provide flexibility for client-specific email servers.
• Developed a dynamic email template resolver using the Liquid templating system, enabling ticket-based dynamic rendering of emails.
• Managed the system’s daily handling of approximately 10,000 emails, alongside leading a team of 2 engineers for maintenance and periodic enhancements.
• Managed multiple projects simultaneously while maintaining strict deadlines and high-quality standards.

Email Ticket Handler

• Developed an email ticket handler using .NET Core APIs, integrating Postmark for email parsing and ticket action automation.
• Designed the parsing and configuration workflow for processing incoming emails.
• System improved ease of use and reduced portal traffic by 30% by enabling email-based support conversations.
• Managed ongoing maintenance and iterative feature enhancements.