Harsh Raj

AI Researcher

Hyderabad, Telangana, India2 yrs 3 mos experience

Most Likely To SwitchAI ML Practitioner

Key Highlights

Expert in developing AI solutions for Snapdragon chipsets.
Proven experience in building LLM-based applications.
Strong foundation in software engineering and web development.

Stackforce AI infers this person is a skilled AI/ML engineer with a focus on developing innovative solutions in Fintech and Web Development.

Contact

Skills

Core Skills

Machine LearningLarge Language Models (llm)Natural Language Processing (nlp)Front-end Development

Other Skills

Python (Programming Language)LangchainTypeScriptRegression TestingAutomationTest AutomationGenerative AI ToolsAI toolsSystem on a Chip (SoC)Computer VisionRapidsAI AgentsGraph DatabasesNeo4jOntologies

About

Passionate ML Engineer dedicated to building innovative AI solutions that address real-world challenges. Currently, I’m part of the AI Software (AISW) team at Qualcomm, where we work on optimizing neural networks to run efficiently on-device using the QNN SDK — enabling powerful AI experiences on Snapdragon chipsets. Previously, I worked at TCS as an ML Engineer, where I built RAG-based LLM solutions and generative AI products for enterprise use cases. I also interned at CIMB AI Labs, contributing to LLM-driven automation tools, and at Hobnobly, where I strengthened my software engineering foundation. Excited by the future of AI and always looking to collaborate on meaningful, high-impact projects.

Experience

2 yrs 3 mos

Total Experience

9 mos

Average Tenure

10 mos

Current Experience

Qualcomm

Engineer - AI/ML

Jul 2025 – Present · 10 mos · Hyderabad, Telangana, India · On-site

Part of AI Software Team - enabling AI and ML models to run efficiently on Snapdragon chipsets.

Python (Programming Language)Machine Learning

Tata consultancy services

Machine Learning Engineer (R&I)

Aug 2024 – Jul 2025 · 11 mos · Hyderabad, Telangana, India · On-site

Built a multi-agent system with dynamic agent routing via LangGraph for context-aware decision flows.
Developed a custom knowledge graph that allowed agents to assess query completeness and intelligently prompt users for missing information.
Integrated MCP for smooth inter-agent communication, with full experiment tracking in MLflow.
Developed a chemical domain QnA chatbot using LangChain to extract insights from research papers.
Implemented a custom solution to retrieve relevant images from research papers, enhancing research.
Worked on Nvidia AI Tech - Nvidia Metropolis, RAPIDS and Omniverse.

LangchainLarge Language Models (LLM)

Cimb

Data Scientist Intern

Mar 2024 – Aug 2024 · 5 mos · Bengaluru, Karnataka, India

Enhanced efficiency by optimizing BANCA codebase, achieving an 8x speed improvement.
Spearheaded a POC initiative for LLM-based text extraction from financial documents of different formats.
Automated AML case narratives generation with Mistral 7b, integrating tabular data, preset rules, and few-shot training to replace manual workflows and enhance efficiency.
Worked on Time series analysis utilzing the Halt Winter methods.

Python (Programming Language)Natural Language Processing (NLP)

Amazon

Mentee - Amazon Science ML Summer School

Sep 2023 – Oct 2023 · 1 mo · Remote

Attended Amazon ML Summer School, guided by industry experts through comprehensive sessions.
Gained in-depth knowledge of advanced machine learning concepts, including Supervised Learning, Deep Neural Networks, Dimensionality Reduction, and Unsupervised Learning.
Explored complex topics such as Probabilistic Graphical Models, Sequential Learning, Causal Inference, and Reinforcement Learning.
Significantly enhanced theoretical understanding of machine learning techniques and their practical applications.

Python (Programming Language)Natural Language Processing (NLP)

Hobnobly - discover anyone's calendar and build communities

Software Development Intern

Jul 2022 – Sep 2022 · 2 mos · Remote

Refactored company's website ensuring improved performance and user experience.
Improved website performance by achieving a remarkable 97% image compression rate through PNG to WebP conversion, resulting in faster loading time.
Deployed the website using EC2 instances and implemented streamlined code management and maintenance by managing branches.
Demonstrated proficiency in CMS platforms and web development by developing and deploying a Blog website using WordPress.

Front-End DevelopmentTypeScript