Jakki Chaitanya
Eswar Rajesh

Pioneered a 13-agent LangGraph DAG that leverages Claude Code to convert legacy codebases into runnable stacks with audits, reports, and zero Claude Max API cost.

AI-powered law enforcement and investigation platform unifies voice, crime intelligence, crime mapping, facial/vehicle deblur, and video narrative, built on Angular + FastAPI.

An AI assistant automating government RFP response generation and resume-to-template population, running locally with an LLM for cost-effective, quick results.

Upload a resume → LLM extracts structured candidate data → fills DOCX reference templates automatically via semantic field mapping. Replaced a 45-minute manual HR task with a fully automated 2-minute workflow.

Turns dashcam and bodycam footage into structured law-enforcement incident reports. Frame-by-frame ViT-GPT2 captions + Whisper audio transcript → Groq LLaMA 3 generates both a plain-English summary and a formal police narrative as structured JSON.

Auto-extracts skills from DOCX/PDF resumes using local Mistral LLM and populates standardised HR skill matrix templates via RAG. Fully local — no candidate data sent to external APIs. UUID-based session storage with rotating log handlers.

Zero human input — Gemini generates SEO-optimised scripts → gTTS voiceover → MoviePy video composition → YouTube Data API auto-publish. Runs on GitHub Actions cron, no server required. Evolved into a modular AI Educational template with dedicated stages for curriculum generation, slides, code demos, animations, and narration audio.

Full-stack recruitment platform with FastAPI backend and TalentCore React Native/Expo candidate app. LLM analyses resume-job fitment with match score, strengths, and gap analysis. Dual-role chat (recruiter + candidate), resume upload, job posting, and SQLAlchemy-backed data layer — designed for local LLM deployment.

Fully automated daily pipeline: fetches trending AI & Tech news from RSS feeds → Gemini 2.5 Flash writes a 45-second script, caption, and hashtags → Microsoft Edge TTS generates voiceover with word-level timing → MoviePy + FFmpeg assembles 3 topic-matched Pexels stock clips into a 1080×1920 vertical video with animated captions → auto-publishes to YouTube Shorts and Instagram Reels. Includes a Gemini-powered auto-fix agent that diagnoses failures and opens a PR automatically.

Chaitanya Eswar Rajesh Jakki's dynamic developer profile, featuring AI/data engineering projects, an interactive career infographic with detailed breakdowns, and automated updates.

AI-powered Streamlit app using FastAPI, ChromaDB, and local Ollama LLM to parse resumes, fill templates, and answer questions via RAG.

AppNovaAI modernizes legacy code into runnable, audited stacks line-by-line using Claude Code, DAG-driven RAG agents, guardrails, and comprehensive reports.

Ollama-powered full-stack app with Angular, optional Streamlit UI, and FastAPI backend generates and packages starter apps, data, and validation.

Automated GitHub Actions bot publishes daily Bhagavad Gita verses in Hindi, Telugu, English as Reels/Shorts using Gemini AI, Google TTS, Pillow, FFmpeg, Cloudinary.