loading
All projects
2025LLM-driven educational workflows saving ₹60+ LakhsActively maintained
ScriptIQ & Broadcast Automation
An automated LLM scripting pipeline producing 1,000+ educational scripts/yr and slashing channel broadcast creation from 30 mins to 2–5 seconds.
₹60L+cost savings
2–5screation time
200+channels automated
At scale, generating hundreds of high-quality educational broadcast scripts and video outlines manually creates significant bottlenecks and costs. ScriptIQ was architected as an automated, multi-stage LLM production pipeline to eliminate these inefficiencies.
The pipeline breaks down complex curriculum topics into structured knowledge graphs, drafts narrative scripts with verified pedagogy constraints, and conducts automated self-consistency audits using specialized prompting techniques. The system saved over ₹60 Lakhs in operational overhead within its first year while generating 1,000+ complete scripts.
In addition, I designed and deployed a YouTube broadcast automation engine supporting 200+ active channels. By leveraging asynchronous parallel execution, asset pre-computation, and intelligent caching layers, video package setup times plunged from 30 minutes down to 2–5 seconds per broadcast.
highlights
- Automated 1,000+ educational scripts/year saving ₹60+ Lakhs in operational costs
- Multi-stage prompt-engineered verification ensuring curriculum fidelity and tone compliance
- Cut broadcast creation time from 30 minutes to 2–5 seconds across 200+ YouTube channels
- Parallel task queuing and Redis caching yielding 6x throughput gains under peak load
- Python
- OpenAI
- LangChain
- FastAPI
- Redis
- Docker
- GCP
- YouTube API