Jul 2024 – Mar 2025 · 8 mos · Dubai, United Arab Emirates · Remote

• Engineer No. 2
• Incredible teammates from OpenAI, NeuralLink, Nvidia, Apple, Meta, Google.
• Core Tech Stack & Innovations:
• 1. Web3 AI Decentralised Tech Stack – Built a robust and scalable AI-first decentralized infra.
• 2. Fastest Routing Intelligence Engine – Developed a routing model that outperforms Meta LLAMA 3.1 70B, Qwen2.5 70B, and all open-source LLMs using router-based optimization.
• Lead research efforts for the routing engine, surpassing benchmarks on BBH, MMLU, MUSR, GPQR.
• Distributed computing via Ray Serve with multiple hosted H100 GPUs, deploying multiple LLMs for internal use cases.
• Designed Ocean – a Perplexity.ai-style system but 20x faster for real-time AI inference.
• 3. Voice Agent – Built using LiveKit, enabling real-time AI-powered conversational agents.
• 4. Marketing Agent Platform – Reward-based system for content engagement, integrating blockchain & on-chain analysis for O Token distribution and anomaly detection.
• 5. High-Speed Compute Infrastructure – Deployed on world’s fastest AI supercomputer Cerebras WSE-3, alongside high-performance H100 clusters.
• 6. Advanced LLM Evaluation & Security
• Developed in-house Auto Eval LLM Judge to combat hallucinations & jailbreak attempts.
• Built highly optimized Knowledge Base RAG with custom chunking logic.
• In-house Jailbreak Guard for LLMs to ensure safe AI interactions.
• 7. Computer Vision & NLP – Implemented data collection, large dataset management, and qualitative data analysis.
• 8. On-Chain Analysis & Anomaly Detection – Built blockchain-powered monitoring for token distribution and fraud detection.
• Tech Stack:
• LLMs & AI Models – GPT,LLaMA, Mistral,Qwen, custom routing models.
• Distributed Computing – Ray Serve, Mlflow, AWS, Docker.
• Generative AI – Advanced Reinforcement Learning models for optimal AI training.
• Cloud & Infrastructure – SageMaker, HF integration.

Apr 2024 – Aug 2024 · 4 mos · Remote

Jan 2023 – Apr 2024 · 1 yr 3 mos · Remote

• Engineer No. 6
• 1. Generative Models
• 2. Large Data Pipelines , MLOPs
• 3. Recommendation Engine
• 4. Predictive Modeling
• 5. AWS Tech stack

Nov 2020 – Dec 2022 · 2 yrs 1 mo · United States

• Engineer no. 1
• Founding Data Scientist and Core team member.
• 1. Build the data foundation for the company, aggregating over 900 Million - 1.2 Billion rows of airfare data on BigQuery, GCP.
• 2. Build and Scaled Arbitrage Model, taking arbitrage profit in ticket booking.
• 3. Got patent for the arbitrage model
• 4. Predicitve Booking, Automatic Booking, Recommendation Engine

Aug 2020 – Oct 2020 · 2 mos · London, England, United Kingdom

• 1. Scrapped over 3,00,000 instances of car data from gum tree, craiglist, motors.co.uk , autotrader
• 2. Deployed Car Valuation estimation from different attributes based on car model, mileage, brand, transmission, etc
• 3. Deployed Stock and revenue prediction (like peak.ai)

Aug 2020 – Sep 2020 · 1 mo · Israel

• 1. Scrapped ~96,000 UPC products from online stores for food ingredients
• 2. Built Automatic Data Normalization from Different source deployed on AWS S3 bucket
• 3. Designed and deployed Machine Learning models for nutrient detection like Gluten-Non Gluten, GMO-NonGMO, Fat freeFat models.
• 4. Model precision and recall rate over ~88%

Jun 2020 – Jul 2020 · 1 mo

• Peer Reviewer for the Machine Learning and Deep Learning papers in IMU, EVISE.

May 2019 – Nov 2020 · 1 yr 6 mos · India · Remote

• Acquired by SkillArbitrage
• Engineer no. 2
• Founding Data Scientist.
• Started with the core team of 3, build the initial course curriculum and projects and MVP. The company did fairly well from initial students from 10 to 5000 during my tenure of work.

May 2019 – Sep 2019 · 4 mos

• Gesture Recognition , Opticial Flow, Deep Learning , FPGAs

Apr 2019 – Nov 2020 · 1 yr 7 mos

• Top Talent, 100% Job Score, 16+ International clients.