Mohd Arsh

Backend Engineer + AI Builder

I build the backend
that makes AI work.

Based in Mumbai. Building multi-tenant SaaS, LLM-powered pipelines, and automation systems. Currently shipping Extractly.

Mumbai, India · arshhansari27@gmail.com

Selected work

MULTI-TENANT SAAS · EXTRACTLY

CV parsing pipeline that handles 500-file batches without breaking a sweat.

Extractly automated CV parsing dashboard

PRODUCT DESIGN + BACKEND · GHAR GIFT

SaaS for Indian PG owners — tenant management, billing, and WhatsApp alerts.

RAG + SEMANTIC SEARCH · NEXTREAD

Semantic book recommender — LangChain, Chroma, HuggingFace, Gradio.

NextRead semantic book recommender

AUTOMATION + AI · BLOTATO / N8N

Social media automation — Telegram approval loop, per-platform retry, 70-node workflow.

Twitter · LinkedIn

Building

Learning by shipping.

Extractly HRs upload CVs, define custom fields, get structured Excel/CSV exports. Multi-tenant with per-client field schemas, Celery chord batching (100 files × 10/task), GPT-4o + Google Vision extraction, S3 direct upload, openpyxl export at merge layer.

Django 5 + DRF Celery + Redis Gemini + Vision Pydantic v2 Next.js PostgreSQL

Ghar Gift SaaS for Indian PG/hostel owners. Owner → Property → Tenant model, EdDSA auth (Better Auth), WhatsApp notifications via Whapi, Cloudinary media, Celery background jobs. Feature-complete and shipped.

Marketplace & SaaS Django 5 + DRF Better Auth Whapi Next.js Cloudinary

NextRead Semantic book recommender. Describe a book you loved, get 5 embeddings-matched suggestions from a vector store of 7,000+ books. First real RAG project — where I learned that chunking strategy matters more than model choice.

LangChain Chroma HuggingFace Gradio Python

Contact

Open to interesting problems and collaborations.

arshhansari27@gmail.com

Mohd Arsh Ansari