Aparna Pradhan

A full-stack web and React Native developer specializing in AI integration with TensorFlow.js and TFLite, as well as LLM API integration. Experienced in building AI agents, retrieval-augmented generat

india, west bengal Joined on Jan 3, 2025 https://main--aparnapradhanportfolio.netlify.app/

Cover image for The $20 Billion Strategic Warning Shot: Why NVIDIA Fused the LPU into the CUDA Empire

Aparna Pradhan

Dec 27

The $20 Billion Strategic Warning Shot: Why NVIDIA Fused the LPU into the CUDA Empire

#inference #cuda #groq #nvidia

4 min read

Cover image for Clone Your CTO: The Architecture of an 'AI Twin' (DSPy + Unsloth)

Aparna Pradhan

Dec 26

Clone Your CTO: The Architecture of an 'AI Twin' (DSPy + Unsloth)

#ai #llm #rag #agenticai

3 min read

Aparna Pradhan

Dec 25

🚀 GLM 4.7 : Is the era of "expensive-only" SOTA models ending?

#llm #ai #glm #zai

2 min read

Aparna Pradhan

Dec 25

The Perfect Extraction: Unlocking Unstructured Data with Docling + LangExtract 🚀

#docling #langextract #ai #rag

3 min read

Cover image for 🤯 Stop Paying Cloud Bills: How Transformers.js & WASM Shifts RAG Compute to the Browser (Client-Side AI)

Aparna Pradhan

Dec 25

🤯 Stop Paying Cloud Bills: How Transformers.js & WASM Shifts RAG Compute to the Browser (Client-Side AI)

4 min read

Aparna Pradhan

Dec 25

The Research: MiniMax M2.1 (The "Linear" Revolution)

#llm #openai #rag #ai

3 min read

Cover image for Why We Replaced Our Orchestrator with a 'Regex' Switch

Aparna Pradhan

Dec 11

Why We Replaced Our Orchestrator with a 'Regex' Switch

#regex #litellm #llm #orchastration

4 min read

Cover image for Why LLMs Fall Short: Why Large Language Models Aren't Ideal for AI Agent Applications

Aparna Pradhan

Jan 3

Why LLMs Fall Short: Why Large Language Models Aren't Ideal for AI Agent Applications

#ai #rag #llm

3 min read

Forem

Aparna Pradhan

Badges

Writing Debut

Skills/Languages

Currently learning

Currently hacking on

Available for

The $20 Billion Strategic Warning Shot: Why NVIDIA Fused the LPU into the CUDA Empire

Clone Your CTO: The Architecture of an 'AI Twin' (DSPy + Unsloth)

🚀 GLM 4.7 : Is the era of "expensive-only" SOTA models ending?

The Perfect Extraction: Unlocking Unstructured Data with Docling + LangExtract 🚀

🤯 Stop Paying Cloud Bills: How Transformers.js & WASM Shifts RAG Compute to the Browser (Client-Side AI)

The Research: MiniMax M2.1 (The "Linear" Revolution)

Why We Replaced Our Orchestrator with a 'Regex' Switch

Why LLMs Fall Short: Why Large Language Models Aren't Ideal for AI Agent Applications