Engineering Intelligent Systems.
Focusing on
Hi, I am Chintan. I design and scale production LLM and RAG systems, from hybrid retrieval and agentic orchestration to the distributed infrastructure that runs it at multi-tenant scale.
Work Experience
A chronological timeline of engineering leadership, AI systems architecture, and production deployments.
AI Engineer
RIB SoftwareJul 2025 to PresentNashik, MaharashtraDesigning and scaling a production RAG platform: distributed backend architecture, multilingual retrieval pipelines, and AI infrastructure serving 50+ enterprise tenants.
Key Systems Delivered & Accomplishments
- Productionized a construction Bill-of-Quantities RAG platform from proof-of-concept to production scale, enabling 10x faster estimate generation at 99.9% uptime with sub-second p95 query latency.
- Architected the platform's distributed backend, FastAPI async APIs, Celery task orchestration (50,000+ jobs/week), Redis, and Azure Service Bus, processing 100,000+ multilingual (English/German/European) BOQ items through an end-to-end RAG pipeline.
- Designed conversational AI and RAG solutions using MCP-based orchestration across multi-source knowledge bases, improving contextual retrieval accuracy by 30%; built advanced RAG pipelines incorporating HyDE, RAG Fusion, and re-ranking, increasing retrieval precision by 25%.
- Built and published a reusable AI/ML platform infrastructure library covering pluggable LLM/embedding provider factories and Celery worker orchestration, adoption across AI microservices eliminated 70%+ code duplication.
- Scaled the platform from 0 to 50+ enterprise tenants in 12 months with zero-downtime deployments, handling 10,000+ API requests/day, deployed on Azure Kubernetes Service with Helm and Azure DevOps CI/CD, backed by pytest coverage at 80%+.
FastAPICeleryRedisAzure Service BusLangChainRAGMCPAzure AI SearchKubernetesHelmDockerJul 2025 to PresentNashik, MaharashtraSoftware Developer
Winjit Technologies Pvt. Ltd.Jun 2024 to Jul 2025Nashik, MaharashtraBuilt OCR and NLP document ingestion pipelines and machine learning models for classification and fraud detection, deployed as RESTful services across UAT, test, and production environments.
Key Systems Delivered & Accomplishments
- Developed OCR and NLP ingestion pipelines in Python using Google Vision API, automating processing of 7M+ financial documents at 90% accuracy.
- Trained and evaluated ML models for document classification, intent detection, and fraud analysis through feature engineering and exploratory data analysis, achieving 97% fraud-detection accuracy.
- Deployed RESTful ML services across UAT, test, and production environments with structured logging and end-to-end traceability, sustaining 95% detection accuracy across all stages.
- Streamlined document processing workflows via a performance-tuning module, reducing per-transaction turnaround time by 5 seconds and increasing overall operational throughput.
PythonOCRGoogle Vision APINLPMachine LearningREST APIsJun 2024 to Jul 2025Nashik, Maharashtra
Featured Projects
Curated repositories spanning autonomous agent architectures, RAG indexing pipelines, and inference optimization tools.
DL-with-PyTorch
Deep learning experiments in PyTorch spanning ANN, CNN, and RNN architectures, including a Fashion-MNIST image classifier.
SpaceX-Falcon-9-first-stage-Landing-Prediction
End-to-end data science project predicting SpaceX Falcon 9 first-stage landing success: API/web-scraping data collection, SQL-based EDA, geospatial analysis with Folium, predictive modeling, and an interactive Dash dashboard.
Movie-Recommendation-System
Movie recommendation engine built on the TMDB 5000 movie dataset in a Jupyter notebook.
Movie-Booking-System
Backend movie ticket booking system built in Go with hold booking feature.
Technical Skills & AI Stack
Comprehensive overview of frameworks, models, vector stores, and engineering principles used across production deployments.
AI & LLM Engineering
RAG architecture, agentic orchestration, and applied deep learning, taken from proof-of-concept to production.
Backend & Infrastructure
Async APIs, distributed task orchestration, and the messaging layer that runs production AI workloads.
Cloud & DevOps
Deployment, scaling, and CI/CD for containerized AI services on Microsoft Azure.
Data & Quality
Storage, observability, and test coverage across production AI and data pipelines.
Languages
Core programming languages used day to day and across past projects.