Available for Projects

Harshil Lakhani, AI full-stack developer and automation engineer: I buildfull-stack systemsthat automate business operations.

From SaaS products and internal tools to AI workflows and integrations, I turn manual processes into reliable software teams can use every day.

0%

Rendering frames

HARSHIL LAKHANI

AI FULL-STACK&AUTOMATION ENG
Services & Solutions

What I Build

End-to-end software and AI automations tailored for growing businesses and modern products.

workflow-automation-pipeline.ts
WORKFLOW AUTOMATION

I link your everyday business tools (CRMs, email, spreadsheets, APIs) with intelligent AI agents that handle data entry, customer routing, and background operations 24/7 without manual effort.

Example Live Automation Flow⚡ Speed: 42ms
1. Trigger
New Lead / Order
──────▶AI Analyzes
2. Action
Sync CRM & Slack
Key Deliverables You Receive:
End-to-End n8n Workflows
Autonomous AI Task Bots
Automated Web Scraping
Instant Slack & CRM Alerts
Zero Manual InterventionProduction Ready

Selected Work

A showcase of problems solved and systems built.

Signature Project
Enterprise
NDA Protected

Support Space | 3PL Operations Platform

Built a role-aware operations platform that brings seller management, order investigations, omnichannel support, AI-assisted replies, tasks, credits, audit trails, analytics, and warehouse attendance into one internal system. The interface is designed around high-volume 3PL work: automate routine steps, surface exceptions clearly, and keep operators in control of every consequential action.

Next.jsFirebaseGmail APIFreshdeskChannelDock
Read the 3PL operations platform case study
Project Views
16
Product Pages
97
Backend Routes
5
Operational Workspaces
tap or hover a project view to explore the deck
One operating system for modern fulfilment teams
Platform overview
One operating system for modern fulfilment teams
Seller command center
One operating system for modern fulfilment teams
Global order search
One operating system for modern fulfilment teams
Support workspace
One operating system for modern fulfilment teams
AI-assisted replies
One operating system for modern fulfilment teams
Auditable operations
One operating system for modern fulfilment teams
Warehouse control
One operating system for modern fulfilment teams
Business impact
Interactive Impact Analysis

Manual Friction vs. Automated Scale

Drag the slider to compare standard manual operations with custom-engineered autonomous systems.

WITH AUTOMATED SYSTEM
99.8% Effort Reduced
< 200ms Execution

Real-time webhooks, asynchronous queues, and automated multi-agent triage without human bottlenecks.

0% Human Error

Structured schema validation, automatic escalation rules, and auditable error logs.

$0.02 / Operation

Replaces thousands in repetitive manual labor with high-throughput cloud infrastructure.

24/7 Continuous Uptime

Self-healing task retries, live monitoring, and synchronized multi-platform data flow.

Architecture: Next.js + Python + LangGraph + WebSocketsStatus: Fully Scalable
BEFORE (MANUAL PROCESS)
High Overhead
4.5 Hours / Day

Operators wasting hours copy-pasting data between spreadsheets, email inboxes, and portals.

14% Mistake Rate

Missed carrier deadlines, lost customer tickets, and delayed shipment resolutions.

$3,500+ Monthly Cost

Burning payroll budget on repetitive administrative tasks that should be automated.

Fragmented Tools

5 different browser tabs open with zero unified audit trails or consolidated reporting.

Workflow: Manual Data Entry & Email TriageStatus: Bottlenecked

More Projects

3PL Email Automation | Carrier Investigations & Seller Portal screenshot 1
3PL Email Automation | Carrier Investigations & Seller Portal screenshot 2
3PL Email Automation | Carrier Investigations & Seller Portal screenshot 3
3PL Email Automation | Carrier Investigations & Seller Portal screenshot 6
Enterprise

3PL Email Automation | Carrier Investigations & Seller Portal

Built a logistics investigation SaaS that turns a fragmented support process into a structured seller portal and internal operations workflow. Sellers can review shipped orders, raise carrier queries, upload evidence, track case status, and receive updates synchronized from carrier email replies and Freshdesk tickets. The implementation spans portal flows, backend logic, Firestore data handling, and integrations with ChannelDock, Gmail, Google Drive, and browser-based Freshdesk automation.

Browser AutomationPythonFirebaseGoogle APIsNode.js
Warehouse TV Dashboard | Live Staffing & Task Control screenshot 1
Warehouse TV Dashboard | Live Staffing & Task Control screenshot 2
Warehouse TV Dashboard | Live Staffing & Task Control screenshot 3
Warehouse TV Dashboard | Live Staffing & Task Control screenshot 8
Enterprise

Warehouse TV Dashboard | Live Staffing & Task Control

Built a real-time warehouse operations dashboard for a fulfilment team to manage staffing, assign work across departments, monitor active tasks, review employee history, and export operational reports. It includes warehouse TV views, task timers, attachments, business-hour-aware tracking, and fast workflows for supervisors coordinating day-to-day fulfilment activity.

Next.jsReactFirebaseNode.jsWebSockets
AI Returns Scanner | Label Reading & Returns Intake screenshot 1
AI Returns Scanner | Label Reading & Returns Intake screenshot 2
AI Returns Scanner | Label Reading & Returns Intake screenshot 3
Enterprise

AI Returns Scanner | Label Reading & Returns Intake

Built a returns intake app for a Dutch 3PL that handles around 160 returned parcels a day for more than 340 sellers. Workers photograph the shipping label, Gemini on Vertex AI reads the order number, customer name and tracking code, and the app matches the parcel to its ChannelDock order line. Each item is scanned and marked good or damaged, then the return is booked and the screen shows where the goods go. Announced Bol and Amazon returns are received instead of duplicated, and parcels that cannot be matched safely go to an exceptions shelf.

Next.jsTypeScriptGemini (Vertex AI)ChannelDock APIFirestorePWA
Shift Board | Carrier Cutoff & SLA Planning screenshot 1
Shift Board | Carrier Cutoff & SLA Planning screenshot 2
Shift Board | Carrier Cutoff & SLA Planning screenshot 3
Enterprise

Shift Board | Carrier Cutoff & SLA Planning

Built a live planning board for a Dutch 3PL fulfilment company that compares the open pick backlog with each carrier's cutoff. It reads orders from ChannelDock, sorts non-pickable orders into eight exclusion buckets, tracks pick rates over 15, 30 and 60 minutes, and forecasts the rest of the day from 15-minute weekday curves. Shift leads get picker deficit advice and a what-if simulator. Scheduled jobs send each carrier's SLA email exactly once after its deadline. The board also replaces the daily Excel schedule with a PDF-ready builder and reads rosters from screenshots.

Next.jsTypeScriptFirebaseChannelDock APIGeminiRailway Cron
Sorter Monitor | Parcel Sorter Lane Accuracy screenshot 1
Sorter Monitor | Parcel Sorter Lane Accuracy screenshot 2
Sorter Monitor | Parcel Sorter Lane Accuracy screenshot 3
Enterprise

Sorter Monitor | Parcel Sorter Lane Accuracy

Built a monitoring service for a Dutch 3PL fulfilment company that receives a copy of every parcel sorter scan through a secured webhook. It detects the carrier from 14 barcode rules, compares the expected carrier lane with the lane the parcel actually took, and counts retried scans only once. Counts are kept per hour in Firestore with only a few writes, and the admin panel emails managers the accuracy, every wrong-lane parcel and any unknown barcode formats. Key rotation, an IP allowlist, rate limits and body-size limits protect the only endpoint the sorter supplier can reach.

Node.jsTypeScriptFirebaseGmail APIRailway
Carrier Portal Bot | Claims & Ticket Automation screenshot 1
Carrier Portal Bot | Claims & Ticket Automation screenshot 2
Carrier Portal Bot | Claims & Ticket Automation screenshot 3
Enterprise

Carrier Portal Bot | Claims & Ticket Automation

Built a browser automation service for a Dutch 3PL that works inside a carrier partner's Freshdesk customer portal, which has no API. The bot opens parcel investigation and claim tickets with the carrier, claim reason and sales invoice attached, posts follow-up comments, reads replies and closes resolved tickets. It runs Playwright through Camoufox behind a FastAPI service, reuses saved sessions, and allows one browser at a time. Ticket lists and change checks run over plain HTTP in about half a second, so the browser starts only when a ticket has actually changed.

PythonFastAPIPlaywrightCamoufoxDocker
PropTrader AVM | AI Property Intelligence Platform screenshot 1
PropTrader AVM | AI Property Intelligence Platform screenshot 2
PropTrader AVM | AI Property Intelligence Platform screenshot 3
PropTrader AVM | AI Property Intelligence Platform screenshot 6
Web

PropTrader AVM | AI Property Intelligence Platform

Built a full-stack AI property intelligence platform that turns an address and property specifications into a structured valuation workflow. It combines automated valuation, comparable-property context, portfolio wealth tracking, macroeconomic scenario simulation, property news, and downloadable appraisal reports in one investor-facing product.

Next.jsTypeScriptFastifySupabaseVertex AI
TerraScope AI | Property Feasibility Platform screenshot 1
TerraScope AI | Property Feasibility Platform screenshot 2
TerraScope AI | Property Feasibility Platform screenshot 3
TerraScope AI | Property Feasibility Platform screenshot 5
Web

TerraScope AI | Property Feasibility Platform

Built a property feasibility platform that takes a residential site from uploaded documents to a costed report. Zoning certificates, title deeds and stand schedules go through a RAG extraction pipeline where each chunk is verified or rejected by a person. Verified constraints keep their source snippet and confidence score, then feed a design brief and several concepts scored on privacy, views, buildability, cost and compliance. Costing is deterministic, shown in rand as a range with a confidence percentage, and it lists missing inputs instead of guessing them. The feasibility report is assembled from the verified data, the chosen concept and the cost estimate.

Next.jsTypeScriptFastAPISupabaseRAG

Experience

Built Through Real Product Work

Full-stack development across operational software, automation systems, backend services, and client-facing products.

Freelance Full-Stack Automation Engineer

Self-Employed

April 2026 - Present

Building full-stack web applications, internal operational tools, and autonomous AI workflow pipelines that eliminate manual bottlenecks.

Full-Stack Automation Engineer

Nextbase Solution

September 2025 - March 2026

Engineered scalable web systems, automated SEO publishing pipelines, and multi-service API integrations with robust database workflows.

Full Stack Developer

Coderbabu Infotech

January 2023 - August 2025

Developed responsive web and mobile applications with Node.js backends, managing database architectures and full product lifecycles.

Writing

Engineering Notes

Practical breakdowns of automation workflows, AI systems, and the decisions behind them.

Visit the n8n & AI automation blog
Instant ROI Estimator

Calculate Your Automation ROI

See how much time and operational budget your team can recover by replacing manual workflows with custom automation software.

Repetitive Hours Spent / Week20 hrs/wk
5 hrs40 hrs (1 Full-Time Staff)80 hrs
Average Hourly Employee Cost$35/hr
$15/hr$50/hr$150/hr
Estimated Business Recovery ~85% Recoverable

Annual Operational Savings

$30,940/ year

Monthly Budget Saved

$2,578/mo

Hours Saved / Year

884 hrs

Start a Project

Have a product, internal tool, or workflow that needs a reliable technical solution? Tell me what you are trying to improve.

Tell Me About Your Project

Available for projects

Let's Build Something Useful

I work with teams that need full-stack products, internal tools, AI workflows, backend systems, or reliable integrations.

India · IST (UTC+5:30)
Response: < 2 hrs

Direct Line (Call / WhatsApp)

+91 8140010069