GenAI developer task completed#9
Open
ShivaKumarKaranam2 wants to merge 1 commit intodivamtech:mainfrom
Open
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
GenAI Developer Assignment Submission - Shiva Kumar Karanam
Video Summary - LinkedIn Generator - DOCX Bulk Engine - Character Video Pipeline
Overview
This PR contains my complete submission for the GenAI Developer assignment. It demonstrates end-to-end system design thinking across multiple GenAI use cases, focusing on architectural clarity, structured prompting, scalability, and production-readiness.
The solutions emphasize practical trade-offs (privacy, cost, quality, bulk processing), structured JSON outputs, multimodal reasoning, and modular pipeline design aligned with real-world deployment considerations.
Scope of Work
Problem 1 --- Video-to-Notes System
Compared three architectures: SaaS-based, hybrid cloud LLM, and fully offline.
Included Mermaid diagrams, trade-off analysis, cost considerations, and structured JSON schema for summarization.
Addressed privacy, scalability, and review-loop design.
Problem 2 --- Zero-Shot LinkedIn Post Generator
Designed a single-call prompt producing three stylistically distinct drafts.
Strict JSON schema enforcement for structured output.
Persona alignment and formatting consistency prioritized.
Problem 3 --- Smart DOCX Template → Bulk DOCX/PDF Engine
Multimodal GenAI-based field detection from template files.
Automated bulk document generation via Excel/Sheets.
Fidelity-focused rendering using docxtpl + LibreOffice PDF conversion.
Error handling, reporting, and scalability considerations included.
Problem 4 --- 5-Min Character Video Series Pipeline
Modular pipeline: script → storyboard → assets → audio → render.
Character consistency via structured "Bible" injection.
Forward-looking 2026 tool stack integration (Runway Gen-4.5, Kling 2.6+, Gemini 2.5/3, Qwen models, etc.).
Emphasis on repeatability and production workflows.
Design Principles Demonstrated
Structured JSON outputs for reliability
Zero-shot prompt robustness
Clear architecture trade-offs and recommendations
Human-in-the-loop review stages
Bulk processing and error management
Practical production constraints (cost, latency, privacy)
Status
All four problems fully addressed
Trade-offs and recommendations clearly articulated
Prompts and schemas validated for structural reliability
Content self-contained and professionally structured
I welcome feedback on architectural decisions, prompt robustness, or areas where deeper technical validation would strengthen the proposal.