Sep 2025 – Jan 2026 · 4 mos · Bengaluru, Karnataka, India · Remote

• Client : Wadhwani Foundation
• Leading the National Jobs Portal for Wadhwani Foundation, architecting an agentic AI platform with RAG pipelines and multi-agent workflows for job discovery, skill-matching, and career guidance.

Oct 2024 – Aug 2025 · 10 mos · Bengaluru, Karnataka, India · On-site

• Bootstrapped the company with ₹50L personal investment; closed FY 2024–25 at ₹1.8Cr revenue
• Pioneered “Set Selling” model – bundling premium venues with fully produced décor sets; transformed 4 properties into profitable
• wedding destinations
• Onboarded premium venues including Vivanta, Merian, and other top Guwahati properties
• Served 300+ clients in FY 2024–25
• Grew social media following to 10K+ across platforms
• Represented the company at IMC and multiple investor presentations
• Secured media coverage; Times of India to feature Assamese wedding at flagship venue on 29 Nov
• Built a 7-member core team; scaled operations to handle large-scale, high-budget weddings

Jun 2020 – Sep 2024 · 4 yrs 3 mos · Bengaluru, Karnataka, India · On-site

• 1. Architected Jio HES+ Copilot: GPT-4 (instruction tuned + CoT exemplars) for query normalization & routing, on-prem Llama-3.1-70B (vLLM) for composition, Qwen2.5-Coder for MongoDB aggregation & plotting.
• 2. RAG Layer: Weaviate vector store with bge-m3 embeddings, bge-reranker, PyMuPDF for SOPs/tariffs/policies.
• 3. Agentic Execution: Multi-agent routing across billing, NILM, feeder ops, PQ, geo; JSON guardrails; Serper API for live updates.
• 4. Codegen Bridge: Auto-generation & execution of Mongo pipelines + plots in sandbox.
• 5. Impact: Unified 20+ dashboards into one Copilot; sub-2.5s p95 latency; officer engagement +300%.
• 6. Jio HES+ Analytics: Scaled to process 1M+ smart-meter data points daily across thousands of DTs/LMVs.
• Billing & Revenue: Thru rate, billing efficiency, exceptional/corrected bills, due-date deviation, collection efficiency, revenue across LMV/HV, segmentation of late payers.
• Consumption: Avg/peak/off-peak usage, load vs temp/humidity, peer benchmarking, high/low bill alerts, COVID-era shifts.
• Reliability & PQ: SAIFI, SAIDI, CAIDI, MAIDI/MAIFI, PF analysis & penalties, voltage scoring, PQI, BQI.
• Event & Tamper: Categorization of outage, PQ & tamper events (magnetic, reverse, bypass, cover open); hotspot maps for raids.
• Asset Health: Over-utilized DTs, stress monitoring, poor-performing asset scoring, composite Health Index.
• Predictive ML/DL: Short-term load forecasting (ARIMA/XGBoost/LSTM/TFT), DT stress forecasting, theft detection (Isolation Forest + XGBoost), high-bill alerts.
• Synthetic Theft Data: DCGAN-based; 10,000+ cases generated from 2,000 tampered meters at Jamnagar.
• Impact: Improved billing efficiency, reduced AT&C losses, enabled proactive asset planning, customer nudges, and leadership adoption.
• 7. Leadership & Culture: Managed 4 DS GETs, ran Kaggle-style leaderboards, weekly paper reviews, and encouraged juniors to challenge champion models.
• 8. Delivery: 13 champion models promoted into production across domains.

Oct 2018 – Apr 2020 · 1 yr 6 mos · Mumbai, Maharashtra

• Played a key role in developing state-of-the-art 3D vision systems for segmentation, classification, detection, 3D bounding boxes,
• and 6D pose estimation from LiDAR point clouds, deployed in the flagship product Ajnalnes.
• Led R&D on motion and performance capture, creating deep learning solutions for human and hand pose estimation, deployed as
• a scalable web service on Google Cloud.
• Advanced research on 3D avatar creation from a single RGB image, contributing to innovation in immersive and virtual environments.
• Optimized and deployed computation-heavy deep learning algorithms on resource-constrained edge devices (Intel Neural Com-
• pute Stick, Qualcomm 845 dev kit).
• Research Contributions: Co-authored 1 conference paper and contributed to 3 patents during tenure

Jun 2018 – Oct 2018 · 4 mos · Kharagpur Area, India

• Developed vision algorithms for quality assessment of agricultural products, which includes quality assessment of rice grains and
• wheat grain using deep learning algorithms.
• De3veloped the core AI algorithms for TRAGNEXT which automates faulty and manual Fine Leaf Count process in Tea manufacturing.
• Helped over 1400 tea cultivators of different TATA TEA estates to get fairly paid because of this innovation.
• Deployed in Nvidia Jetson.
• Drone image analysis for surveillance. A POC was done and presented to the Lt. Gen. of Indian Army DGIS

Nov 2017 – Feb 2018 · 3 mos · Portland, Oregon Area

• Worked on technical bench-marking of different machine learning frameworks based on the training time and memory usage of
• various machine learning algorithm.
• Performance of optimization techniques for data with varying sparsity were compared

May 2016 – Aug 2016 · 3 mos · Bengaluru Area, India

• Classification of 3-D urban LiDAR point cloud.
• Relevant features from the LiDAR point cloud was extracted which was classified with random forest in a conditional random field
• framework.
• Proposed a novel semi-supervised technique for classification of LiDAR point cloud.
• Worked on clustering of the LiDar point cloud for the development of a visualization tool.