Jul 2024 – Nov 2025 · 1 yr 4 mos · San Francisco, California, United States · On-site

• ➤ Developed a cutting-edge memory solution for AI agents to allow highly personalized interactions through adaptive context retention.
• ➤ Engineered a memory-centric LLM pipeline, increasing answer quality by 26% compared to OpenAI and reducing token and latency costs by 90%.
• ➤ Developed scalable memory retrieval pipelines and established evaluations to measure performance across add and search methods.
• ➤ Implemented a Neo4j-based graph-memory for efficient entity storage and retrieval, boosting context awareness and response accuracy.

Jan 2024 – Jul 2024 · 6 mos · London Area, United Kingdom · Remote

• ➤ Improved image quality globally and locally through selective detail enhancement, achieving a 23% increase in customer satisfaction.
• ➤ Developed a Siamese network for detecting similar HDR bracket images, achieving artifact-free merging with 0.82 Silhouette Score.
• ➤ Engineered an image censoring system with 91% mAP@0.5, masking sensitive data (license plates, people faces, etc.) to ensure privacy.

Aug 2022 – Dec 2023 · 1 yr 4 mos · Marina del Rey, California, United States · On-site

• ➤ Utilized prompt engineering and memory-based approaches in large language models to enhance text-based game playing.
• ➤ Designed reinforcement learning agents with enhanced memory, language grounding, and object affordances. [Published at K-Cap 2023][Best Student Paper Award]
• ➤ Enhanced visual question answering system by implementing visual cropping methods to focus specific regions in images. [Published at NeurIPS 2023]
• ➤ Developed a multimodal approach to generate recipes from food images using attention-based vision and language models. [Published at WACV 2024]

Jul 2020 – Aug 2022 · 2 yrs 1 mo · Gurugram, Haryana, India

• ➤ Utilized analytical and technical expertise to deliver insights and suggestions for business progress.
• ➤ Achieved 92% precision on fraud engine by leveraging historical and session-level attributes to block fake leads on platform.
• ➤ Developed a computer vision pipeline to create 3D floor plans from spherical projections, saving 400 work hours/month.
• ➤ Automated image auditing process, saving $40k/year and increased online traffic by 12%.
• ➤ Devised an image classifier with 93% precision and 89% recall, labeling real-estate images in real-time at 6 FPS on CPU.
• ➤ Published two research papers in top International Conferences: ECML’22 and CODS-COMAD’23.
• My contribution resulted in savings and an increase in online traffic. My company recognized my efforts by honoring me with the Exemplary Performer for the Year 2020-21 Award. In addition, I have been awarded the Spotlight Award twice for each Ownership and Excellence.

Jan 2020 – Jul 2020 · 6 mos · Gurugram, Haryana, India

• ➤ Improved image quality (contrast, brightness, sharpness, blur) resulting in 8% increase in CTR on social media campaigns.
• ➤ Developed a No-Reference Image Quality Ranking tool to select the most visually appealing real-estate property images.
• ➤ Launched a service that extracts real-estate entity tags from WhatsApp group messages with 87% f1-score.