AI Engineer · Bengaluru, IN

Nishaanth
K N

I build AI systems that reason and predict in production — not just in notebooks. LLMs, RAG pipelines, real-time data, and the infrastructure that keeps them honest.

View Projects Get in touch
scroll
About

Who I am

Computer Science graduate from Amrita Vishwa Vidyapeetham, now a Software Engineer at Tarka Labs, Bengaluru. Previously a Data & AI Engineer at Bosch Global Software Technologies.

My focus sits at the intersection of LLMs, real-time data pipelines, and cloud infrastructure. I care about systems that handle edge cases gracefully, not just happy paths.

Outside work: solo travel (Singapore, Vietnam), long-form blogging, and six days a week in the gym.

Python / ML95%
LLMs / RAG90%
SQL / Data Engineering85%
Azure Cloud (AKS, Event Hubs)80%
PyTorch / Deep Learning78%
Experience

Where I've worked

Software Engineer
Tarka Labs · Bengaluru
May 2026 — Present
Building production-grade AI features and backend systems. Working across LLM integrations, API design, and cloud infrastructure to ship AI products that work reliably at scale.
Python LLMs Azure REST APIs
AI / ML Engineer
Bosch Global Software Technologies · Hyderabad
Aug 2025 — Apr 2026
Built and deployed NLP-powered internal tools and RAG systems serving cross-functional engineering teams. Designed LLM-backed pipelines on Azure infrastructure and created Streamlit-based interfaces for model-driven analytics.
RAG LLMs NLP Azure Streamlit
Machine Learning Intern
Bosch · Bengaluru
Feb 2024 — Jul 2024
Worked on NLP and predictive modelling prototypes. Contributed to model evaluation pipelines and data preprocessing workflows using PyTorch and scikit-learn.
PyTorch NLP Python scikit-learn
Projects

Things I've built

01
BERT-Based Intent Classifier
Fine-tuned a BERT model for multi-class intent classification on domain-specific enterprise text. Achieved significant accuracy gains over baseline TF-IDF approaches, with a FastAPI inference layer for production serving.
BERT HuggingFace PyTorch FastAPI Python
02
RAG Pipeline — Internal Knowledge Base
Production RAG system for querying internal engineering documentation. Built retrieval, chunking, and re-ranking layers on top of Azure infrastructure with a Streamlit interface.
Python LangChain Azure Streamlit FAISS
03
Sentence Similarity Engine
Built a semantic search and deduplication engine using Sentence-BERT embeddings. Used for clustering support tickets and surfacing duplicate issues across large unstructured text corpora at scale.
Sentence-BERT FAISS PyTorch Python scikit-learn
04
NLP Analytics Dashboard
Interactive Streamlit dashboard surfacing NLP-derived insights — sentiment trends, topic clusters, and anomaly signals — from unstructured data sources at Bosch.
Streamlit PyTorch NLP HuggingFace

Let's talk

Open to interesting AI/ML engineering roles, freelance projects, or just a good conversation about building things that actually work.