Harish Kumar N P
AI Intern
About
Hey, thanks for checking me out!I'm Harish, currently a fourth-year ECE student at IIT Guwahati. I recently worked as an AI Intern at LTID, a Japan-based startup, which exposed me to the international startup ecosystem. It gave me the opportunity to work with real-world datasets for a major company in the Philippines, apply the AI knowledge I had learned to practical projects, and improve an existing credit scoring model.Little summary about me:-Switched from non-tech domain to AI/ML at the start of my 3rd year and have made considerable progress while continuing to learn new technologies.-Secured 3 internship offers for Summer 2026 — 2 off-campus(one revoked due to business issues) + 1 on-campus.-Currently a Knight on LeetCode.-Built genuine projects inspired from some ideas i have seen and approached with a new method-which can be scaled to real time products with addition of few features .-Focused on roles that go deep into applied AI/ML — model development, GenAI systems, and research-driven problem solving.My areas of interest include Machine Learning, Deep Learning, Generative AI, Signal Processing, Natural Language Processing, Data Analytics, and Product Management.Open to full-time roles starting June 2027, and open to a 6-month remote internship starting January 2027 (leading to FTE conversion). Reach me at nphk1609@gmail.com/harish.p@iitg.ac.in
India
Chennai
Information Technology & Services
Teamwork, lang chain, Google Gemini, Adversarial Training, Docker, FastAPI, resnet, self attention, rffi fingerprint, AI Agents, Multi-agent Systems, groq, Git, LightGBM, Feature Engineering, Credit Risk Modelling, Data Cleaning, LangGraph, Data Structures, Computer Networking
Experience
AI Intern
Long Term Industrial Development(LTID)
Built and validated a credit risk model to predict auto loan defaults for JACCS Philippines. Built a LightGBM fraud detection pipeline with target encoding, achieving 0.92 AUC and 0.81 PR-AUC on 3,500+ unseen contracts. Tuned the model using Neyman-Pearson optimization to maximize recall while keeping the false positive rate under 3.2%, reaching 81% recall. Found and fixed data leakage issues in the feature engineering pipeline to identify flaws that affected their reliability. Benchmarked the production model against 10 alternative architectures — ensemble stacking, DART/GOSS boosting, custom asymmetric loss functions — confirming near-optimal convergence via systematic negative-result validation. Received a Letter of Recommendation for extraordinary performance and bringing down false negatives by 29% at a similar FPR.
Education
Indian Institute of Technology, Guwahati
Electronics and Communication Engineering
AA grade in OOPs and Data Structures, Digital Signal Processing, Signals and Systems, Semiconductor Devices, MA201( Partial derivatives and Complex Analysis), Control Systems, CH101, CH110 lab, PH110 lab, EE102 lab
Harish Kumar N P's Contact Information
Phone
Find the Right Leads
Find Verified Contact Data
What LeadContact does well
Find verified emails, phone numbers, and decision-makers with 98% accuracy.
Find Leads
Find the right people by company, role, industry, location, and more.
925M+ professional profiles
Find Emails
Access verified email addresses for your target contacts.
657M+ emails
Find Phone Numbers
Get cross-validated phone data from multiple top sources.
239M+ phone numbers
More Accurate. Lower Cost.
Find contact data in 1 tool with 98% accuracy
LeadContact integrates leading enrichment tools to deliver more accurate contact data—without paying for each one.
Great conversations start with the right contact.
It’s time to find yours.