AI Engineer focused on building production-grade agentic AI systems, RAG pipelines, and scalable backend infrastructure.

github

Ashish Ashish  • Peerlist

Built a production-grade multi-agent AI workforce platform for autonomous financial services automation using LangGraph, FastAPI, ChromaDB, and Groq-powered Llama 3.3 inference. Engineered autonomous Sales, Ops, and KYC agents with: • LangGraph orchestration pipelines • RAG retrieval over compliance documents • AML/sanctions verification workflows • Airtable CRM integration • Gmail & Slack automation • FastAPI backend services Designed the system around scalable agent workflows, structured orchestration, and real-world operational automation.

OmniForce AI Workforce — Multi-Agent AI Syste

Designed and deployed a multi-agent AI system for automated patent analysis, executing novelty detection, infringement mapping, and Freedom-to-Operate (FTO) workflows. Built a high-performance RAG pipeline over 7,000+ patent embeddings using FAISS and sentence-transformers, enabling source-backed reasoning over large-scale patent data. Architected a 6-agent asynchronous pipeline (Search, Filter, Novelty, Claim Mapping, Legal Reasoning, Citation) using Python asyncio, reducing latency from ~30s to <3s (12x speedup). Integrated Groq-accelerated LLaMA 3.1 inference with strict JSON schema validation to eliminate hallucinations and ensure deterministic outputs. Designed system scalable to millions of patent documents with autonomous retrieval + reasoning execution. Tech Stack: Python, FastAPI, FAISS, LLaMA 3.1, Groq, Streamlit, asyncio, Docker, Sentence-Transformers

PatentPilot-AI

Designed and deployed a scalable LLM-powered pipeline for extracting structured knowledge from large-scale scientific literature. Built a hybrid retrieval + extraction system supporting PDF uploads, ArXiv parsing, and raw text ingestion, generating structured JSON outputs from unstructured research documents. Processed large-scale scientific datasets (200M+ papers via APIs and retrieval systems), enabling automated information extraction and summarization workflows. Reduced end-to-end latency from ~45s to <4s by migrating from local inference (Ollama/Qwen) to Groq-accelerated LLaMA 3.1 with optimized prompt execution. Engineered a reliability layer with retry logic, schema validation, and normalization pipelines to ensure deterministic, failure-resilient outputs. Designed modular architecture integrating LLM pipelines with semantic retrieval for scalable and production-ready deployment. Tech Stack: Python, LLaMA 3.1 (Groq), Streamlit, Sentence-Transformers, Semantic Scholar API, Async Processing