available immediately · open to full-time rolesI hand your team’s busywork to AI agents.
Aman Kumar Chhari, AI agent & automation engineer. I map how your team really works, then build agents and workflows that take the repeated steps off their plate. Checked by code, approved by people.
> right now automating:orders and paymentsshop adminquote checksdata collectionreview checks
- AI agents
- Claude API
- Fetch.ai uAgents
- WhatsApp flows
- Webhooks
- Razorpay
- 01 Trigger
Buyer types a shopping list
- 02 AI agents
Ask open shops nearby for quotes
- 03 Check
Code blocks hidden fees and wrong totals
- 04 Done
Order placed, owner taps YES
// proof, not promises
4
agent roles working one order
streetwise
7
automated checks per product scan
scancart.ai
27
eval cases guarding the shop AI
closeby
15 min
until unpaid orders cancel themselves
closeby
Hi, I’m Aman.
I start by mapping how a team really works: who does what, in which app, and which steps they repeat every day. Then I automate those steps with agents and workflows, so people stop copying, checking and replying by hand. Since 2023 I've built client products from first brief to handover.
Recent work: Streetwise, CloseBy, ScanCart.ai, Orderly and Stack.
Hackathon winner (Reimagine Web Dev Hackathon, Component Making Hackathon), with national-level finishes at Hacksagon 2025, HackGSon Hackathon and Smart India Hackathon.
B.Tech, Computer Science & Engineering, Jaypee University of Engineering & Technology, Guna (2022 – 2026).
What is AI automation?
AI automation means connecting a language model to the tools a business already uses, like WhatsApp, payments, sheets and databases, so routine jobs run on their own. A message becomes an order. A payment updates the books. A shopping list becomes checked quotes from shops nearby. People only step in where a decision is needed.
How I ship it
A demo that works once is easy. A workflow a business can trust every day takes more care. This is how I approach it:
Start from the real process
I map the steps a team already does, in the apps they already use, before I automate any of them.
Guardrails on every action
Every model call passes one gateway with auth, rate limits, a kill switch and a daily spend cap. Output is checked before anything is saved.
People stay in control
The AI drafts stock imports, replies and restock ideas. Nothing is saved until the owner approves it.
Measured time saved
I count the manual steps a workflow removes. Eval cases test the AI on misspellings, Hindi aliases and junk input before any change ships.
Automations I’ve shipped.
// 2026 · Finding open shops nearbyStreetwiseFour AI agents that find shops nearby, collect quotes, block hidden fees and place the order▸ closeby/
// 2026 · Stock entry from photos and voice listsCloseByLocal shop platform for Guna, with AI tools that take routine work off shop owners▸ scancart/
// 2026 · Reading and judging reviewsScanCart.aiSeven automated checks on a product listing, returned as one counterfeit-risk verdict▸ orderly/
// 2026 · Price and stock repliesOrderlyWhatsApp questions answered from the live catalogue, so kirana owners stop replying by hand// 2026
Stack
Automated daily review: highlights saved from any app come back each morning
- Nx
- Ts
- Cp
- +5
Experience & wins.
Freelance AI & Automation Developer
Independent · Remote · Feb 2023 – Present
- ›Map how a client's team works, find the steps they repeat, and build tools that do those steps for them.
- ›Build AI workflows on the Claude API: tool calling, drafts that people approve, and logs for every run.
- ›Connect apps to the tools clients already use: WhatsApp, Razorpay payments and webhooks, Supabase and Postgres.
- ›Own scoping, timelines, client communication and handover docs.
AI Data & Evaluation Contributor
OneForma & Outlier · Remote · Oct 2026 – Present
- ›Rate and review AI model outputs against written quality guidelines.
Web Development Intern
Zidio Development · Remote · Aug – Sep 2024
- ›Built full-stack features with React, Node.js, Express and MongoDB, and helped with deployment and testing.
- Finalist (600+ teams), Hacksagon 2025, ABV-IIITM Gwalior, as team MANADcertificate
- Selected pitch, Rajasthan AI Builders Pitch-a-Thon 2026 (National Conference on e-Governance), Jaipur, with ScanCart.aicertificate
- National Finalist, HackGSon Hackathon
- Semi-Finalist, Smart India Hackathon
- Winner, Reimagine Web Dev Hackathon
- Winner, Component Making Hackathon
- 7th Rank, Skill-A-Thon 2024
My automation toolkit.
- Claude API80%
- Node.js80%
- TypeScript85%
- Postgres75%
- Next.js90%
- React90%
AI agents
- Claude API & SDK
- Fetch.ai uAgents
- Prompt design
- Tool calling
- Multi-agent pipelines
- LLM evaluation
Automation
- WhatsApp flows
- Webhooks
- Razorpay
- Scheduled jobs
- Background workers
- Agent-to-agent messages
Integrations
- REST APIs
- OpenStreetMap
- Supabase
- Firebase
- PostgreSQL
- Clerk
- Chrome extensions
Infra & languages
- Python
- TypeScript
- Docker
- VPS
- Git
- Vercel
I automate
- Orders and payments
- Shop admin
- Quote checks
- Data collection
- Review checks
How I build.
- 01Map before I build.I write down how the work happens today before I automate any of it.
- 02Every action has a check.Code checks what an agent wants to do before it happens, and every model call is logged.
- 03Retries are safe.Idempotent webhooks and money stored in paise, so a repeat never charges twice.
- 04Handover is part of the job.Plain-language docs, so the team can run the workflow without me.
What I’m looking for.
Roles
- AI Agent & Automation Engineer
- Forward Deployed Engineer
- AI Automation Engineer
- Type
- Full-time, open to contract
- Level
- Entry level / SDE-1 equivalent
- Location
- Open to relocate
- Start
- Immediately
Teams I’d fit
AI automation agencies · SaaS operations teams · Startups automating their back office