ML Engineer Intern
Computer Science (IoT) student with a minor in AI and internship experience building computer vision pipelines, AI agents, RAG systems, and scalable backend infrastructure.
Matrice AI
Dec – Feb 2026 · Remote, India
Engineered core computer vision pipelines for customer analytics, reducing latency by 15% for real-time streams. Fine-tuned custom YOLO models (JointBDOE) and accelerated inference with TensorRT. Prototyped a Smart People Counter using YOLOv11s, achieving 0.92 mAP on high-density mall datasets.
EngageOS.ai
May – Sept 2025 · Remote, India
Led end-to-end development of an AI Agent-as-a-Service platform for automated multi-channel customer engagement. Architected FastAPI and LangGraph infrastructure handling 1,000+ concurrent agent requests with 99.9% uptime and less than 3 seconds median response time. Implemented RAG pipelines using Qdrant VectorDB and Redis caching, and optimized proactive and reactive agent workflows to reduce operational cost.
Drytis Inc.
2 mos
Matrice AI
4 mos · San Francisco, California, United States
EngageOS
4 mos
Worked on building an AI Agent-as-a-Service platform that helps brands engage with communities across channels like Twitter, Telegram, and Web Widgets. Designed and implemented the LangGraph-based Core AI Engine with reactive + proactive flows. Integrated RAG with Qdrant Vector DB for knowledge-grounded responses. Developed features like language detection, sentiment & relevance filtering, proactive posting strategies, and configurable agent behaviors. Worked with FastAPI, OpenAI/Claude APIs, Redis and Qdrant Cloud to productionize the system. Focused on clean code, DRY practices, and production-ready architecture for scalable AI agents.
B.Tech, Computer Science (IoT), Minor in AI
July 2022 – May 2026
CGPA: 8.00
Bachelor of Technology - BTech, Computer Engineering
Jul 2022 - Jun 2026
Built a real-time voice and text translation platform supporting 20 languages with translated audio playback. Developed a multimodal pipeline using Groq Whisper large-v3 for speech-to-text, LLaMA 3.3 70B for medical translation, and Edge-TTS for speech synthesis. Implemented role-aware translations and structured clinical summaries covering symptoms, diagnoses, and follow-ups.
Built a LangGraph-based multi-agent AI platform with six specialized agents for data ingestion, preprocessing, model training, and evaluation. Implemented conditional workflow routing, LLM-powered model selection and hyperparameter tuning, and a Streamlit human-in-the-loop approval system.
Architected a distributed real-time monitoring system with Python agents and a central FastAPI backend for OS metrics ingestion and analysis. Developed an Isolation Forest anomaly detection engine that reduced false-positive alerts by 30–40% compared with static thresholding, and containerized the stack with Docker Compose.