Independent · Delhi NCR

Automation & Web Developer

Your team’s manual work, built into systems that run themselves.

Automation, internal tools and websites for small teams. I’ve delivered for clients in Germany, France, the UK, India and the UAE.

Tell me what you do by hand
Bhomik Verma

01 · About

About

I build websites, automations, and internal tools: digital systems that help businesses work, operate, and grow better.

I started in QA, testing AI products, which taught me how software breaks. That pulled me into automation: first n8n, then deeper, because I loved the impact the work had. The real job turned out to be understanding the business: going deep into how a team actually operates before changing anything.

Over time I moved to coded automation: it’s more reliable, and it costs less to run. That’s how I build now: custom code, built to order, and at handover you get the full repository.

Before working independently I worked full-time building automation systems for two overseas companies and three other clients. One of them, Metis Labs in Berlin, wrote me a letter of recommendation and offered to act as a reference.

Built for

  • Infinium Associates
  • Goethe Care Institut
  • NexGen Architects
  • SkillCircle
  • White Marketing

02 · Selected work

Client work, and the systems behind it.

50+ projects shipped. These are the ones I kept a record of.

  1. 01

    Infinium Team Tracker

    Internal tool · Infinium Associates, Dubai

    A private tool for Infinium's sales team. Each person logs their calls, meetings and research against a prospect, a day bar shows the hours nobody has accounted for, and a daily summary goes to management by email. The prospect list blocks duplicates, so two people never chase the same company. Front-end in Next.js, data in Supabase.

    Next.jsSupabaseInternal tool

    Infinium Team Tracker: a team member's day, with logged time, activities and the end-of-day report
  2. 02

    Infinium Associates

    Client website · Recruitment firm, Dubai

    A site built for a real client and handed over. Front-end in Next.js, with the content structured so I can update it on request. Live in production on their own domain.

    Visit the live site

    Next.jsWebsiteClient

    Infinium Associates website infiniumassociates.com
  3. 03

    Goethe Care Institut

    Client website · Nursing school, Frankfurt

    A bilingual site for a state-recognised nursing school near Frankfurt: German throughout, with a full English version and separate sections for the training programmes, the faculty, the campus and admissions. Front-end in Next.js. Shipped this month and live on their own domain.

    Open the site

    Next.jsBilingualWebsite

    Goethe Care Institut website goethecare.de
  4. 04

    GEO-SEO Asset Generator

    Automation · Metis Labs, Germany

    Enter a domain and the system scrapes it, classifies every page, and generates production-ready schema, llms.txt and robots.txt, validated and zero-warning. What took two to three hours of manual work per domain now takes three to five minutes.

    Watch the walkthrough

    AutomationSchema.orgScraping

    Generated schema validating clean in the Schema.org validator Walkthrough
  5. 05

    Viral Video Architect

    Content pipeline · Next Automation, France

    Turns raw media into finished 9:16 vertical videos: automatic split-screen layout, lip-synced AI avatar, captions, and publishing to TikTok, YouTube and Instagram with per-platform metadata. Production time per clip dropped from an hour to about a minute.

    Watch the walkthrough

    AI avatarVideo pipelinePublishing

    A finished 9:16 video from the pipeline: split screen, burned-in captions, lip-synced avatar Walkthrough
  6. 06

    AI Customer Support Chatbot

    RAG automation · Bookleaf Publishing, New Delhi

    A retrieval-augmented chatbot trained on the company's own FAQs and policies, answering authors 24/7. Over 70% of routine queries handled without a person, 120+ staff hours saved per month, response time down from one to two days to seconds. Complex cases route to a human.

    RAGAISupport

    RAG chatbot workflow

03 · How I work

From first call to handover.

  1. Step 01 · Before we start

    Scope and access agreed in writing

    Before I start, we agree in writing on scope, access and how your data is handled, with an NDA if you need one. Access is limited to what the job needs and revoked when it’s done.

  2. Step 02 · Discovery

    Start from how the work really happens

    Whether it’s a website, an internal tool, a chatbot or an automation, I start with the people who’ll use it: what they do today, where it slows them down, and what it has to plug into. Then I write up what I’ll build and what “done” means, so there are no surprises later.

  3. Step 03 · Build

    You see it working early

    You get a working version early (the first pages of a site, a dashboard on your real data, a chatbot answering your actual questions) and we adjust from there. The tools fit the job: Next.js for websites, custom code for internal tools, chatbots and automations.

  4. Step 04 · Delivery

    Half now, half on delivery

    Half the payment up front, the other half only after you’ve got what I promised: built, working and handed over.

  5. Step 05 · Handover

    You get the repository

    At handover you get the full repository and everything needed to run it. If we stop working together, your code isn’t locked to me.

04 · Contact

Send me a process.

Describe one thing your team still does by hand. I’ll reply within a day with how I’d automate it and what it would cost.

bhomik@bhomikverma.com · LinkedIn · Zoltworks

3D head scan: Lee Perry-Smith (Infinite-Realities), CC BY 3.0