CONNECTING SYSTEMS

The Architect of Impact

Building Generative AI solutions that bridge the digital divide. Specialized in RAG, Low-Resource NLP, and Enterprise Analytics.

CS-DS Final Year / Heritage Institute of Technology
01 / Intelligence Layer

AI SYSTEMS

NLP96%
Multi-Agent91%
RAG Systems88%
LLM Ops84%
Token Attention
buildwithaiagentsthatreasonandactautonomously
Semantic Embeddings
LangGraph
0.96
RAG
0.91
Agents
0.88
FastAPI
0.82
Embeddings
0.79
Qwen3
0.74
Live Agent — Blog Empire Pipeline● LIVE
NEWS INTEL
──►
SCOUT
──►
WRITER
──►
REVISOR
──►
PUBLISHER
──►
DEV.TO / HASHNODE
fetching live news...
02 / Experiments

SELECTED WORK

Hackathon Winner

KRISHAK: Agri AI

First SMS-based AI advisory system in India for feature phones. 3-second response latency.

01
NLP & Search

Multi-Agent RAG

LangChain and FAISS for high-precision retrieval. Reduced query resolution time by 30%.

02
MLOps & Meta-Learning

AutoML Agent

Automated architecture orchestration. Explores 50 neural structures in just 3 minutes.

03
Backend Systems

DocuGen

FastAPI semantic backend automating codebase documentation. 70% database cost reduction.

04
Backend Architecture

Secure Banking

Robust logic utilizing C++/Python for highly secure financial workflow integration.

05
LangGraph · Multi-Agent Pipeline

Blog Empire

Autonomous AI blogging platform — Scout→Writer→Revisor→Publisher pipeline. Real-time news from 3 APIs, self-healing SEO, Telegram bot control, auto-publishes to Dev.to & Hashnode.

06
AI Email Intelligence

EVA

LangGraph agents for automated email classification, summarization & draft generation. 92% accuracy. OAuth 2.0, Fernet encryption, JWT sessions. Human-in-the-loop review workflow. 4.8/5 satisfaction.

07
Legal Intelligence · Production

AgentX (Euron)

Multi-agent legal backend — contract review, risk analysis, compliance checking & redlining. Supabase/Postgres persistence, full auth integration, comprehensive test suite.

08
03 / Principles

HOW WE WORK

01 / INDUSTRIAL

INDUSTRIAL INNOVATION

Gen AI Intern at TCS. Focused on Large Language Model integration and pipeline scalability.

02 / COMMUNITY

COMMUNITY MENTORSHIP

DSA Mentor at Noobuild. Guiding developers through algorithmic optimization and efficient computing.

03 / LOW-LATENCY

LOW-LATENCY AI

Engineered KRISHAK achieving a 3-second response latency across 36+ states.

04 / SEMANTIC

SEMANTIC PRECISION

Utilized LangChain and FAISS, improving query resolution speeds by 30%.

04 / Network

CONTACT & INFO

Identity

I specialize in scalable backends, low-resource Gen AI, and creating impactful technology for the real world. Let's collaborate.

© 2026 ARKAPRABHA BANERJEE. All rights reserved.

Architect of Impact.