Sai Bhava Teja C

About –
Spiti is an AI-powered video wiki for teams. Think of us as an internal YouTube for organizations.

Leadership –

• Led engineering, product and design efforts.
• Launched a few times on Product Hunt and marketplaces such as Zoom, Google Workspace, and Slack – got our first fewpaying customers, and worked closely with them to improve the product.
• Raised pre-seed funding and grew the team size to 5.

Key Individual Contributions –

• Loom-like desktop screen recorder in electron and vue.js, along with a proxy server to push app updates.
• Framework for easy IPC communication across multiple windows in desktop applications (To be Open-sourced).
• Optimized video imports from sources such as Zoom Cloud and Google Drive by file chunking and parallelizing uploads.
• Reduced import time of a 45 minute Zoom meeting recording to less than 15 seconds.
• Created pipelines for video processing - transcoding, transcribing, and indexing for search.
• Created vector embeddings for semantic search.
• Integrated with OpenAI’s GPT-4 and Google’s PaLM to power AI-driven meeting insights.
• Developed Q&A feature akin to ChatGPT on videos.
• Access control lists on videos to secure videos from unauthorized access.
• Integrated with Stripe for all things billing.
• In-app, email, and Slack notifications about user activity and to improve collaboration.
• Technologies used: Nuxt, Vue, NodeJS, Electron, Cloud firestore, Cloud functions, Docker, GCP, AWS, OpenAI, GPT, LangChain, Typesense, Pinecone.

• Improved the Event attendee experience with Chat, polls, and Q&As.
• Developed WhatsApp like Message Inbox for event attendees.
• Technologies used: Java, Spring Boot, React, Postgres, Redis

• Designed and developed a first-of-its-kind Policy Engine for the firm. Enabled policy provisioning via Terraform.
• ~80 Chaos experiments, improving resiliency of several applications.
• Designed and developed Chaos Engineering as a Service, integrated with GitLab’s CI/CD Pipeline.
• Exposed API Gateway plugin latencies as prometheus metrics, charted graphs in Grafana for better visibility.
• Exposed Active health checks as a configuration in terraform resource for services on-boarded onto the API Gateway.
• Technologies used: Java, Guice, Dropwizard, Terraform, Open Policy Agent, Chaos Engineering, MongoDB, React, ElasticSearch, Python, GraphQL, Prometheus, Grafana, Kong.