Case study
DocuQuery & ETMS
AI document search and an operations suite for Yamato Logistics India
Two prototypes for Yamato Logistics India: DocuQuery, an offline AI that reads scanned records and answers questions, followed by ETMS, an expense and operations suite.

01Overview
A considered experience.
Built for real people.
Yamato Logistics India, part of a Japanese logistics group, wanted to digitise years of scanned paperwork and then run its internal operations on one system. The first prototype, DocuQuery, reads scanned PDFs with OCR, sorts them into categories such as accounts, HR, leasing and legal, and lets staff search by meaning or chat with the documents, with every answer citing its page. It runs fully on-premise with a local language model, so no document leaves the office. The second prototype, ETMS, grew from that work into an operations suite: business trips and claims, mileage, advances, departmental requisitions, purchase orders, vendor onboarding, asset checks, a SaaS risk register and approval and finance consoles, demonstrated to the client behind Microsoft sign-in.
Built at True Value Infosoft for Yamato Logistics India.
02Key features & capabilities
Automatic document categories and per-document summaries
Semantic search and chat with page-cited answers
Fully offline: local embeddings and local LLM
ETMS: trips, claims, mileage, advances and requisitions
ETMS: purchase orders, vendor onboarding, asset checks and approvals
03Tech stack
04Technical architecture
The technology behind the experience.
- DocuQuery: single FastAPI service with OCR, chunked embeddings and cosine search
- Streaming answers with source lists
- On-premise by design so records never leave the network
- ETMS: clickable React prototype planned for Azure AD, HRMS and Tally integration
05Explore the product

Let?s work together
Interested in working
together?
I?m open to new projects and collaborations.
Let?s build something great.