Naboraj Sarkar — Developer Portfolio

Naboraj Sarkar (Nishant) — Developer, AI practitioner, and automation engineer.
Location: Siliguri, West Bengal, India 🇮🇳 • Online: @naborajs

I build audit-friendly automations, Telegram/WhatsApp bots with GitHub logging, and AI agents used in real-world deployments. This repository is a public, searchable record of those projects, technical notes, and learning resources (NSCodex).

---

Table of contents

• About — who I am and what I build
• Featured projects — technical summaries and stacks
• Architecture highlights — diagrams and example flows
• Deployments & operations — hosting, monitoring, and reliability
• Skills & tech stack — concrete technologies I use
• How to collaborate — contact, demos, and contribution
• SEO keywords & meta suggestions (for web landing pages)
• License

---

About

I’m a pragmatic builder who prefers reproducible work and measurable outcomes. I publish my experiments publicly so others can learn from the design, failure modes, and operational details. I focus on:

• AI agents and retrieval-based memory systems
• Automation with n8n and custom runners
• Integrations: Telegram, WhatsApp, GitHub-based logging
• Education through NSCodex: project-first, hands-on learning

Quote: “Build in public, document the failures, and make tools that others can reuse.”

---

Recent highlights (quick)

• Life Logger — Telegram bot → captures structured activity, creates commits in a GitHub repo for every entry (audit-ready personal logs).
• AI Pro Sales Assistant — memory-enabled agent for sales/support with admin dashboard and multi-channel integration.
• CODE — experimental AI-powered IDE for rapid scaffolding using multiple LLM providers.
• NS Fin — crypto & market intelligence dashboard (educational, not financial advice).

---

Featured projects — technical summaries

Life Logger — Telegram bot with GitHub integration

• Type: Bot / Integration
• Stack: Python/Node.js, Telegram Bot API, GitHub REST API, webhooks
• Architecture (summary): Telegram webhook → capture service → normalization → GitHub commit API (JSON/MD) → optional static UI on Vercel
• Value: auditable personal logs, versioned notes for habits and compliance

AI Pro Sales Assistant

• Type: AI agent / Automation
• Stack: LLMs (OpenAI/Gemini/Anthropic), retrieval (vector DB), webhook connectors (Telegram/WhatsApp), admin UI (React)
• Features: memory layers, human-in-the-loop overrides, analytics, continuous updates from new conversations

CODE — AI-assisted online IDE (prototype)

• Type: Web app / Productivity tool
• Stack: React frontend, Node/Flask backend, configurable LLM providers
• Notes: supports user-supplied API keys, multi-model prompt orchestration, scaffold generation

NS Fin — Finance & Crypto Intelligence (educational)

• Type: Data aggregation & dashboard
• Stack: Market APIs, data pipelines, visualization, sentiment analysis
• Focus: research and education — not investment advice

---

Architecture highlights — examples

Life Logger flow (example):

1. Telegram bot receives message → webhook posts to capture endpoint
2. Capture service normalizes message to JSON and validates schema
3. Commit generator prepares a content file and calls GitHub REST API to create a commit in a dedicated logging repo
4. Optional: static UI renders logs from the repo via Vercel for quick browsing

Design principles: idempotence for webhooks, small single-purpose services, clear auditability (git commits), and defensive retries.

---

Deployments & operations

• Hosting: Vercel / Netlify for static sites and lightweight front-ends
• Automations: n8n hosted or self-hosted runners for scheduled workflows
• Monitoring: structured logs, Git-backed changes, and simple alerting (email/Telegram)
• Reliability: exponential backoff on external APIs, circuit-breakers for LLM calls, minimal privileged credentials

---

Skills & tech stack

• Languages: Python (primary), JavaScript/Node.js, basic C++
• Web: HTML/CSS, React for small admin UIs
• APIs: REST, webhooks, GitHub API, Telegram Bot API
• Automation: n8n, cron, webhook orchestration
• AI: prompt engineering, retrieval systems, multi-model orchestration
• Deployment: Vercel, Netlify, simple CI patterns

---

How to collaborate

• Quick demo / bot walkthrough: Telegram — @Nishantsarkar10k
• Business/proposal: nishant.ns.business@gmail.com
• Code, issues, PRs: https://github.com/naborajs

If you want a technical walkthrough (architecture diagram, deployment notes, or code pointers), open an issue in the relevant repository and tag me.

---

SEO keywords & meta suggestions (for publishing this page on a website)

• Meta title suggestion: "Naboraj Sarkar — Developer Portfolio · AI Agents · Automation · NSCodex"
• Meta description suggestion: "Naboraj Sarkar (Nishant) — Developer focused on AI agents, Telegram bots with GitHub logging, and automation. Projects: Life Logger, AI Pro Sales Assistant, NSCodex."
• Primary keywords: Naboraj Sarkar, Life Logger Telegram bot, AI agent development, n8n automation, NSCodex, Telegram GitHub integration
• Structured data suggestion (JSON-LD): add Organization/Person schema on your website landing page to improve search visibility

---

Getting started (for contributors)

1. Browse PROJECTS.md to find a project you want to work on.
2. Open an issue describing the improvement or feature.
3. Fork, implement, and open a pull request with a clear description and a short deployment checklist.

Contribution preferences: small, focused PRs, clear tests or manual verification steps, and documentation updates.

---

License

This repository content is provided under the MIT License.

© 2026 Naboraj Sarkar