Skip to content
Abhijeet Maharana, home

Open to contract and full-time remote work

Senior full-stack engineer for web apps and production AI agents

I'm Abhijeet Maharana. I build production AI agents (tool calling, MCP and live voice), e-commerce and payment systems, and fast React and Next.js apps, from the database to the UI. Contract at 20 to 30 hrs/week, or full-time remote with US and EU teams.

// Simplified: nothing is sent unless a person approved this exact content.
const seal = hmac(key, `${item.kind}:${item.id}:${stable(item.approved)}`);
if (!timingSafeEqual(seal, item.seal)) {
  throw new Error("Not approved");
}

await send(item.approved);

In plain words

The agent works on its own and asks before it acts.

tara-agent drafts work around the clock. Anything that leaves the machine waits as an approval card, and a sealed check makes sure only the approved version goes out.

Read the tara-agent project
5 yrs
shipping production React and Node.js, since 2021
40%
faster core app load, vs. the single-bundle build
7
server tools my open-source voice agent calls mid-conversation
240
tests in tara-agent, my open-source AI agent

Experience: 5 years shipping production React and Node.js, at Muvi.com (50,000+ monthly users) and in my current role at an enterprise product company. Full experience

Services

Full-stack, end to end. Most of what startups hire me for falls into three areas.

Projects

Real systems I built, two of them AI agents. Each is written up as the problem, the design and the result.

Independent build, 2025 to 2026

Full-Stack E-Commerce Platform, Checkout to Delivery

A payment can't be applied twice, and missed confirmations reconcile on their own.

Storefront, two payment gateways, subscriptions, referrals, four shipping carriers, an ops dashboard and a Claude-powered WhatsApp support agent. Exactly-once webhooks, Redis locks and queues on AWS.

  • TypeScript
  • Express
  • tRPC
  • Next.js
  • Expo
  • PostgreSQL

Open source (MIT) on GitHub, 2026

Open-Source 24/7 AI Agent with Human Approvals

Runs scheduled work around the clock, and nothing leaves the machine without a person's approval.

A self-hosted agent that drives Claude Code from Slack and cron, with sealed approval cards, exactly-once scheduling and no secrets in the model's reach.

  • TypeScript
  • Node.js
  • Claude Code
  • MCP
  • Slack Bolt
  • SQLite

Open source (MIT) on GitHub, 2026

Real-Time Voice AI Agent with an LLM Grading Engine

One engine runs any spoken evaluation from markdown files: live voice, visible tool calls, and Claude grading against your own reference answers.

An open-source voice agent on Gemini Live that runs structured spoken sessions, calls server tools mid-conversation, and hands the transcript to Claude for scored, typed reports.

  • TypeScript
  • Node.js
  • Express
  • React
  • Gemini Live API
  • Claude API

Enterprise product company (current role), 2023 to present

Enterprise Component Library and UI Performance

One shared library cut frontend development time by 30%, and code splitting made core pages load 40% faster.

A reusable React component library adopted by 5+ enterprise apps, plus lazy loading and code splitting that cut load times by 40%.

  • React
  • TypeScript
  • Code splitting
  • Jest
  • React Testing Library
  • WCAG/ARIA

Writing

Notes from the projects above: what broke, what I chose and why.

  1. · 8 min read

    Running Claude Code 24/7 on a Mini PC, with Approvals in Slack

    How I run a personal AI agent around the clock on one small machine: scheduled jobs, Slack threads, approval cards before anything leaves the box, and no secrets in the model's reach.

    • AI agents
    • Claude Code
    • Node.js
    • Automation
  2. · 9 min read

    Building a Real-Time Voice AI Agent with Gemini Live and Node.js

    How the parts of an open-source voice agent fit together: 16 kHz and 24 kHz AudioWorklets over a WebSocket to Gemini Live, tool calls run on a Node.js server, and Claude grading the call into a Zod-checked report.

    • AI agents
    • Voice AI
    • Node.js
    • TypeScript

All posts

Working with me from India

The usual worries with a remote hire are overlap, communication and reliability. This is how I handle each one.

Pick a time that suits you

The intro call is 30 minutes on your product and where I can help first. I reply to every message within 24 hours.

Book a 30-min intro

How a contract works

For startups that need senior hands without a full-time hire.

  1. Step 1

    Discovery call

    30 minutes on your product, your backlog, and where I can help first.

  2. Step 2

    Paid starter sprint

    1 to 2 weeks on a real ticket, so you judge the work, not a pitch.

  3. Step 3

    Monthly retainer

    20 to 30 hrs/week as a B2B contractor, with direct ticket ownership.

  • B2B contract, invoiced monthly in USD or EUR
  • Paid by bank transfer (Wise) or through your platform (Deel, Remote and similar)
  • I sign your NDA and IP assignment; all code is yours
  • Rates on the intro call, once I understand the scope

FAQ

What hours do you work?

I set my hours around your team's core hours and fix them in advance: up to 4 hours of overlap with US East mornings, and most of the UK and EU workday.

How do contracts and payments work?

A B2B contract, invoiced monthly in USD or EUR, paid by bank transfer through Wise or through your contractor platform, such as Deel or Remote.

Who owns the code?

You do. I sign your NDA and IP assignment before I start.

How do you communicate?

A written daily update in Slack or your tracker, small pull requests, a short demo every week, and calls for decisions.

Can we start small?

Yes. We start with a paid 1 to 2 week sprint on a real ticket, before any retainer.

Do you handle EU data carefully?

I work only inside your systems, with access you grant and can revoke, and I'm happy to sign your DPA.

Are you open to full-time?

Yes, for remote senior engineer roles with US or EU teams. My notice period is 30 days, and it is negotiable.

What are your rates?

I share them on the intro call, once I understand the scope.

Have a feature, a backlog or a role to fill?

Tell me what you are building. If it is a fit, we start with a short paid sprint.