Skip to content
Devendra Jangiddevendra.pro

What changed, and exactly how

Client names stay private, but nothing else does. Each one shows what was broken, what I actually did, and what measurably changed.

Pharma distributionNorth India16 weeks

Illustrative example — replace with a real engagement before launch

Pharma distributor: 11-day month-end close cut to 2 days

A four-branch pharma distributor ran inventory in one system, billing in another and accounts in spreadsheets. Batch and expiry tracking existed only in a storekeeper's memory. We consolidated onto a single ERP with batch-level traceability and automated the reconciliation that three people were doing by hand.

ERPNextPostgreSQLCustom batch/FEFO rulesAutomated MIS reporting
Read the full case study
11 → 2 days
Month-end close
62%
Reduction in expiry write-offs
Real time
Cross-branch stock visibility
3 FTE
Redeployed from reconciliation to sales support
D2C / e-commerceIndia, pan-national9 weeks

Illustrative example — replace with a real engagement before launch

D2C brand: overselling eliminated across five sales channels

Selling on its own website plus four marketplaces, the brand was overselling roughly 4% of orders and eating cancellation penalties and rating damage. We built a single inventory truth with near-real-time propagation to every channel, plus automated settlement reconciliation.

Node.js middlewareNeon PostgresMarketplace APIsWebhook queue with retries
Read the full case study
4% → 0.1%
Oversell rate
< 30 sec
Stock sync latency across channels
₹ recovered
Short-paid marketplace fees identified in month one
Per-channel
True margin visibility, weekly
ManufacturingWestern India11 weeks

Illustrative example — replace with a real engagement before launch

Manufacturer: 1,900 hours a year of manual work removed

Production planning ran on a whiteboard, job cards were paper, and quality data was typed into Excel twice. We digitised the shop floor flow and automated the reporting layer above it, without replacing the ERP the company had already invested in.

Next.js shop-floor appExisting ERP APIsPostgreSQLScheduled reporting
Read the full case study
1,900 hrs/yr
Manual effort eliminated
2 days → live
Owner MIS availability
18%
Reduction in rework after defect root-cause fixes
₹0
Spent replacing the existing ERP
Retail / aggregationTier-2 cities, IndiaPilot 8 weeks, platform 14 weeks

Illustrative example — replace with a real engagement before launch

Retail network: 60 independent stores buying as one

Sixty independent store owners were each negotiating alone and paying distributor rates. We designed the aggregation model, proved it manually with a twelve-store pilot, then built the shared ordering and settlement platform that now runs the network.

Next.js platformNeon PostgresPayment gateway integrationRole-based partner portal
Read the full case study
9–14%
Improvement in procurement pricing
60
Stores onboarded within two quarters
T+2
Automated settlement cycle
Zero
Settlement disputes escalated past the governance process
Voice AI & edge hardwareGoogle Cloud + on-premise deviceOngoing build

Personal engineering build

A voice assistant that runs on hardware I own

A working voice assistant with no cloud speech bill and no audio leaving my control. A Raspberry Pi with a microphone and speaker talks to an automation hub on a cloud VM, which does the listening, thinking and speaking entirely with open-source models.

Whisper (STT)Piper (TTS)Local LLMRaspberry PiGoogle Cloud VMDocker
Read the full case study
₹0
Per-minute speech API cost
On-premise
Audio never leaves the network
One VM
Fixed, predictable running cost
Extensible
New devices added without server changes
Voice AI & edge hardwareLocal networkPrototype to working system

Personal engineering build

Voice control across several rooms at once

Several small devices, each with a microphone and speaker, all streaming audio to one server. The server works out who said what, matches it against known commands even when the words are not exact, and triggers the right response on the right device.

Faster-WhisperWebRTC VADFastAPIWebSocketsRaspberry PiPython
Read the full case study
Multi-device
Several endpoints on one server
Real time
Response begins as speech ends
Tolerant
Fuzzy matching handles imperfect speech
Low cost
Runs on single-board computers
Data engineering & analyticsIndiaOngoing, running daily

Client engagement — anonymised

Scheduled analytics that catch losses nobody was looking for

A set of scheduled analytics jobs that read the operational database every day, look for patterns a person would never spot by eye, and email the right manager when something looks wrong — while there is still time to do something about it.

RPostgreSQLSQLAutomated emailScheduled jobs
Read the full case study
Daily
Checks run automatically, unattended
Days, not months
Time to spot a problem
One source
Everyone works from the same numbers
Zero manual
Effort to produce the exception reports
In their words

What clients say afterwards

“We had been quoted twice as much by two implementation partners, both of whom wanted to start configuring in week one. Devendra spent two weeks understanding how we actually work before recommending anything. That is why it went live on time.”
Managing Director — Pharmaceutical distribution group
“He did not try to sell us a new ERP. He made the one we had already paid for actually work, and then automated the reporting on top of it.”
Director, Operations — Engineering components manufacturer
“The training was the part I underestimated. Six months later my team was running phase two themselves without calling anyone.”
Founder — Multi-city services business
“We found short payments from marketplaces in the first month of reconciliation that more than covered the entire project cost.”
Founder — D2C home care brand
“The pilot on spreadsheets is what convinced sixty sceptical store owners. He proved the model worked before asking anyone to pay for software.”
Founding member — Independent retail collective
“Straight answers, including the ones I did not want to hear. He told us to delay the project by a quarter and fix our master data first. He was right.”
CEO — Agri-processing company
Talk about your situation
CallWhatsAppEnquire