Nishchal PR profile photo 1 on Tal

Nishchal PR

AI - Building knowledge graph infra & LLM pipelines · 0-​>1 High Agency

Bengaluru

2+ years of work experience

Naples Ventures
Feb 20265 mosPresent

Bengaluru

Architecting the core AI infrastructure layer for a B2B investment Analysis startup, enabling high-throughput transformation of unstructured financial corpora into schema-aligned, graph-native, queryable representations for downstream analytical workflows. Engineering a hierarchical, agent-mediated validation architecture, where successive LLM-driven and rule-constrained agents perform schema enforcement, cross-entity consistency resolution, temporal coherence checks, and relational integrity validation, with iterative correction loops and feedback Orchestrating heterogeneous LLM workloads across LLaMA-based models and locally deployed Qwen 3.5 Coder, leveraging model specialization for extraction, validation, and control tasks, while enabling deterministic offline evaluation, prompt instrumentation, and reproducible testing environments Conducting large-scale empirical evaluations across thousands of document shards, analyzing system behavior across dimensions of extraction fidelity, validation convergence, latency distribution, and failure mode clustering, and iteratively refining system robustness and throughput characteristics
The National Institute Of Engineering, Mysore
Mar 20256 mosAug 2025

Mysore, Karnataka, India · Hybrid

Designed and implemented a Retrieval-Augmented Generation (RAG) pipeline as a major academic project to improve factual accuracy in LLM responses Built an end-to-end workflow including document ingestion, text chunking, embedding generation, vector-based retrieval, and context-aware response generation Integrated retrieval results into the LLM prompt to ground outputs in external knowledge and reduce hallucinations Evaluated system behavior across different queries to analyze relevance, response consistency, and failure cases Developed a modular pipeline architecture to allow easy experimentation with embeddings, retrieval strategies, and prompt formats Gained hands-on experience with applied LLM systems, trade-offs between generation quality and retrieval relevance, and limitations of RAG-based approaches
The National Institute Of Engineering, Mysore
Sep 20246 mosFeb 2025

Mysore, Karnataka, India · Hybrid

Designed and implemented an object detection system as part of an undergraduate minor project using a YOLO-based architecture Performed dataset preparation including image annotation, preprocessing, and train–validation splitting Trained and fine-tuned the detection model to improve precision and recall under limited computational resources Implemented inference logic for detecting target objects from images/video streams using Python and OpenCV Evaluated model performance using standard metrics and analyzed false positives and missed detections Gained practical exposure to the full computer vision workflow: data → model training → inference → evaluation
Nopixel Studios Pty Ltd
May 20244 mosAug 2024

Mysuru, Karnataka, India

Handled real-time decision-making, conflict resolution, and trust & safety operations in coordination with platform leadership. Coordinated with admins, moderators, and content teams to manage peak-traffic events, resolve conflicts, and maintain service continuity under real-time pressure.
Nopixel Studios Pty Ltd
Jan 20244 mosApr 2024

Mysore, Karnataka, India

Assisted in setting up AI-supported moderation workflows for an online community Used automation tools to help manage content filtering and moderation tasks Supported day-to-day community operations through basic AI-driven processes
GDSC NIE
Apr 202310 mosJan 2024

Mysore, Karnataka, India

Collaborated with cross-functional student teams to plan and deliver tech events, improving outreach, engagement, and hands-on learning outcomes Applied real-world tools and workflows (Python, ML fundamentals, APIs, Git, deployment basics) Gained experience in ownership, execution under constraints, and translating abstract concepts into working implementations.

Nishchal’s AI Stack

langchain
langchain
langgraph
langgraph
mcp
mcp
n8n
n8n

Education

The National Institute Of Engineering, Mysore

Bachelor of Engineering - BE, Electronics and Communications Engineering

The National Institute Of Engineering, Mysore

20224 yrs2026

Join Nishchal and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Nishchal PR on Tal

Headline

AI - Building knowledge graph infra & LLM pipelines · 0-​>1 High Agency

About

AI Engineer focused on Graph RAG, agentic systems, and production-grade LLM infrastructure. Experienced in building graph-grounded retrieval pipelines, multi-hop reasoning systems, hallucination mitigation architectures, and scalable AI workflows under real-world constraints.

Location

Bengaluru

AI stack

langchain

langgraph

mcp

n8n

Work experience

AI Engineering Intern at Naples Ventures, Feb 2026 - Present · 5 mos, Bengaluru, Architecting the core AI infrastructure layer for a B2B investment Analysis startup, enabling high-throughput transformation of unstructured financial corpora into schema-aligned, graph-native, queryable representations for downstream analytical workflows. Engineering a hierarchical, agent-mediated validation architecture, where successive LLM-driven and rule-constrained agents perform schema enforcement, cross-entity consistency resolution, temporal coherence checks, and relational integrity validation, with iterative correction loops and feedback Orchestrating heterogeneous LLM workloads across LLaMA-based models and locally deployed Qwen 3.5 Coder, leveraging model specialization for extraction, validation, and control tasks, while enabling deterministic offline evaluation, prompt instrumentation, and reproducible testing environments Conducting large-scale empirical evaluations across thousands of document shards, analyzing system behavior across dimensions of extraction fidelity, validation convergence, latency distribution, and failure mode clustering, and iteratively refining system robustness and throughput characteristics

Retrieval-Augmented Generation Pipeline at The National Institute Of Engineering, Mysore, Mar 2025 - Aug 2025 · 6 mos, Mysore, Karnataka, India · Hybrid, Designed and implemented a Retrieval-Augmented Generation (RAG) pipeline as a major academic project to improve factual accuracy in LLM responses Built an end-to-end workflow including document ingestion, text chunking, embedding generation, vector-based retrieval, and context-aware response generation Integrated retrieval results into the LLM prompt to ground outputs in external knowledge and reduce hallucinations Evaluated system behavior across different queries to analyze relevance, response consistency, and failure cases Developed a modular pipeline architecture to allow easy experimentation with embeddings, retrieval strategies, and prompt formats Gained hands-on experience with applied LLM systems, trade-offs between generation quality and retrieval relevance, and limitations of RAG-based approaches

Object Detection Model Training & Deployment at The National Institute Of Engineering, Mysore, Sep 2024 - Feb 2025 · 6 mos, Mysore, Karnataka, India · Hybrid, Designed and implemented an object detection system as part of an undergraduate minor project using a YOLO-based architecture Performed dataset preparation including image annotation, preprocessing, and train–validation splitting Trained and fine-tuned the detection model to improve precision and recall under limited computational resources Implemented inference logic for detecting target objects from images/video streams using Python and OpenCV Evaluated model performance using standard metrics and analyzed false positives and missed detections Gained practical exposure to the full computer vision workflow: data → model training → inference → evaluation

AI-Augmented Community & Moderation Systems Manager at Nopixel Studios Pty Ltd, May 2024 - Aug 2024 · 4 mos, Mysuru, Karnataka, India, Handled real-time decision-making, conflict resolution, and trust & safety operations in coordination with platform leadership. Coordinated with admins, moderators, and content teams to manage peak-traffic events, resolve conflicts, and maintain service continuity under real-time pressure.

Applied AI Systems Contributor at Nopixel Studios Pty Ltd, Jan 2024 - Apr 2024 · 4 mos, Mysore, Karnataka, India, Assisted in setting up AI-supported moderation workflows for an online community Used automation tools to help manage content filtering and moderation tasks Supported day-to-day community operations through basic AI-driven processes

Core Team Member at GDSC NIE, Apr 2023 - Jan 2024 · 10 mos, Mysore, Karnataka, India, Collaborated with cross-functional student teams to plan and deliver tech events, improving outreach, engagement, and hands-on learning outcomes Applied real-world tools and workflows (Python, ML fundamentals, APIs, Git, deployment basics) Gained experience in ownership, execution under constraints, and translating abstract concepts into working implementations.

Education

Bachelor of Engineering - BE, Electronics and Communications Engineering, The National Institute Of Engineering, Mysore