Mohammad Shokoohi-Yekta

Building a Production-Ready RAG System for Persian Language Processing I’m excited to share insights from developing an advanced Retrieval-Augmented Generation system specifically engineered for Persian/Farsi document processing, leveraging cutting-edge ML infrastructure and architectural patterns. 🔬 Technical Architecture Highlights: Advanced RAG Implementation: ∙ Self-Corrective Agentic RAG with adaptive query rewriting and iterative refinement ∙ Hybrid search architecture combining dense embeddings with SPLADE sparse representations ∙ Multi-agent orchestration enabling context-aware document interaction and general conversation modes Production Infrastructure: ∙ NVIDIA Triton Inference Server deployment on H100 GPUs with vLLM backend ∙ Milvus vector database for high-performance similarity search ∙ PostgreSQL for structured data with MinIO object storage for scalable file management ∙ Model ensemble: Qwen2.5-72B-Instruct-AWQ, DeepSeek-OCR, and sentence-transformers ML Pipeline Innovations: ∙ Automated synthetic QA pair generation using LangGraph workflows ∙ Cross-encoder reranking (ms-marco-MiniLM-L6-v2) for precision optimization ∙ Dynamic prompt management system with A/B testing capabilities and versioning ∙ Fine-tuning pipeline with quality-scored synthetic datasets Engineering Principles: ∙ Clean Architecture with dependency injection and separation of concerns ∙ Docker containerization for reproducible deployments ∙ Comprehensive observability with metrics tracking and performance monitoring This project demonstrates how to architect enterprise-grade RAG systems that combine theoretical ML advancements with practical production considerations, particularly for low-resource languages. Open to discussing technical implementation details, architectural decisions, or challenges in deploying LLMs at scale. #MLEngineering #RAG #NLP #LLMs #VectorSearch #ProductionML #AI #DeepLearning #Persian #TritonInference #CleanArchitecture

1

🧬 The concept of Neural DNA (nDNA), initially proposed by Dr. Amitava Das, opens up a new avenue for interpreting LLM behavior. nDNA treats an LLM as something we can sequence (just like DNA sequencing) by reading its internal signals -- its lineage and layer-level changes. It captures a model’s lineage, tuning-induced mutations, and functional expression so we can compare LLM families, spot drift, and explain why similar benchmarks behave differently across a range of LLMs. In practice, nDNA looks for signals such as: (i) layer shifts, (ii) MoE routing patterns, and (iii) controlled probes. It lets us audit change over time and pick models for tasks with evidence. Collaborating with Dr. Amitava Das, and his Pragya group (https://pragyaai.github.io) at the Department of CSIS BITS Pilani Goa Campus, we've been building and testing the concept of nDNA to understand LLM behavior across major job families: - Llama tends to refine mid–to-late layers while early layers remain stable; - Mistral localizes expertise while the early layers remain steady; - Gemma aligns with light-touch parameter changes; - Qwen rewrites more in its middle layers, especially in multilingual scenarios; - DeepSeek adds dialogue capability in mid layers, while preserving base geometry. One of the other distinct advantages of nDNA is the ability to identify model collapse. On the surface you might see a slow loss of creativity and nuance in a model’s behavior. Inside, through the nDNA lens, the model’s internal landscape flattens -- fewer distinct patterns light up -- much like reduced neural diversity in biology. With nDNA, that shift becomes measurable over time, which makes collapse diagnosable and monitorable. Dr. Amitava Das's novel thought process and insistence on clear hypotheses has led to the inception of nDNA as a practical lens for AI -- one that detects drift before deployment, keeps model lineages auditable, and matches models to tasks based on evidence rather than guesswork. Taken together, these nDNA lenses reframe how we interpret LLM behavior. Our other recent works, DPO-Kernels (ACL 2025; https://lnkd.in/gXGQvXwX), Yin-Yang Align (ACL 2025; https://lnkd.in/gWVCxG8B), QuickSilver (https://lnkd.in/gpZQKMmP), and the Counter Turing Test (EMNLP 2023 Outstanding Paper Award; https://lnkd.in/gaVHeXgJ), share a similar first-principle approach to LLM refinement/understanding. If nDNA is something you'd like to explore further -- we'll be launching a website for nDNA soon, stay tuned! #nDNA #AI #FoundationModels #ModelInterpretability

282

5 Comments

For those interested, here is the deck I presented: https://lnkd.in/gCRW2urM Hopefully it provides more clarity and useful information on some of the burning questions related to career development.

40

Very interesting and unheard of! This is the first course of Scientific Machine Learning in Oil and Gas that doesn’t call it “AI” or “Artificial Intelligence”. Match point for the scientific side! 💯 SC-15: Introduction to Scientific Machine Learning for Geophysicists Monday, 25 August | 8:00 a.m.–5:00 p.m. | George R. Brown Convention Center • Introduction to PINNs • PINNs for solving forward and inverse problems in Geophysics • Introduction to Neural Operators • Solution of forward and inverse problems using neural operators • Physics-informed Neural Operators • Emerging trends in Scientific Machine Learning #SciML #petroleumEngineering #SPE #oilandgas #PINN #PILM #PhysicsOfData #notAI

14

1 Comment

Quantization in Large Language Models (LLMs) As LLMs scale, intelligence is no longer the bottleneck — memory, latency, and cost are. Quantization turns out to be one of the most powerful (and misunderstood) techniques enabling real-world deployment of large models. In this blog, I break down: • Why LLMs are surprisingly quantization-friendly • INT8 vs INT4 vs NF4 — what actually works in practice • How modern architectures absorb low-bit noise • Where quantization fails and how to design around it Quantization is no longer an optimization step — it’s foundational infrastructure for production LLM systems. Link: https://lnkd.in/gqGje75G #LLMs #AIEngineering #Quantization #MachineLearning #GenAI #MLOps

9

2 Comments

Mapping the "Identity" of Quantum States ⚛️ As a life long learner and Qiskit Advocate from a remote area of Pakistan, I’ve often found that the biggest fun in quantum is to see math LIVE! We need practical, reality-based tools to see the math in action. Today, I’m sharing my latest project: an interactive Eigen-Hunter built with Python and Streamlit. This tool bridges the gap between Linear Algebra and Quantum Logic. It’s one thing to solve an equation on paper; it’s another to see the "Click" when an abstract Matrix aligns perfectly with a physical state vector. What this research tool achieves: 🔹 Geometric Intuition: Visualizing an Eigenvector as a "Straight Path" that preserves its identity through a Quantum Gate. 🔹 Arithmetic Verification: A live(Row X Column) breakdown to demystify the transformation. 🔹 Eigenvalue (lambda) Discovery: Seeing how this "Factor" relates directly to quantum measurement results. 🔹 Research Documentation: A built-in log to track stable basis states for global collaborations. By "hunting" for these stable directions, we aren't just looking at numbers; we are finding the basis states that define how quantum computers actually process information. Check out the code and try the Pauli-X Gate challenge: Can you find the state that refuses to flip? 📂 GitHub Repository: https://lnkd.in/d8_kMM_h #QuantumComputing #Qiskit #Python #LinearAlgebra #OpenSource #QuantumResearch #WomenInStem #EigenFunctions IBM Quantum OQI – Open Quantum Institute QuLearnLabs Miguel Alfonso Martinez Qiskit Olivia Lanes Nathaniel Walker Lee Chang Xin

7