Ashutosh Malaviya Email & Phone Number

London, Gb

Transforming Communication Through AI Solutions 🌐Imagine a world where communication empowers, not disempowers. At Q we are on a mission to realize this vision by creating a user-centric and humanistic software solution dedicated to addressing the communication challenges faced by individuals with cognitive and speech disabilities. Our comprehensive toolkit, fueled by AI models, a community of subject matter experts committed to success, aims to bridge gaps in speech, memory and cognition caused by conditions such as aphasia, dementia, alzheimer and many other reasons.

Currently directing AI and software for a stealth mode startup using transformer models like GPT, Claude and LLama.- Fine-tuning foundation models on proprietary data using methods like PERF and LoRA for 70% higher task accuracy.- Built Question Answer systems using Retrieval Augmented Generation with vector databases like Pinecone for semantic search.- Reduced model size by 60% using quantization and pruning while preserving accuracy.- Optimized latency and costs by deploying models on AWS- Developed MLOps pipeline for explainability using LIME, concept drift detection, bias testing and recursive feedback loops. - Created vector embeddings of customer data using sentence transformers like SentenceBERT.- Built hybrid search combining transformer scoring with efficient vector similarity ranking using cosine and Euclidean distance.

Mountain View, Wy, Us

New York, Ny, Us

Created Movable Ink’s AI and Data Science based product suite Da Vinci. Da Vinci personalizes visual experiences across email and mobile to improve customer engagement and transform marketing by increasing the conversion accuracy. Built a team of ML engineers and data scientists to design, develop and deploy ML Models (ensemble of HGBoost, Bayesian, Clustering), large scale data pipelines, and data analytics for large enterprise clients.Created an AI Roadmap for Da Vinci Artificial Intelligence-based product suite delivering personalized, automated emails for Retail, Travel, and other verticals.Designed and architected image and campaign feature pipelines for personalized email marketing using Multi-layout modeling and message contextualization techniques. Developed computer vision and NLP-based propensity scores for creative recommendations.Successfully migrated the Da Vinci ML Cloud Infrastructure from AWS to GCP utilizing cloud tools - EMR, DataProc, S3, DCS, Redshift, BigQuery, Vertex AI, and Sagemaker.Accomplished the Analytics Due Diligence for acquiring startup Coherent Path in Jan 2022 and integrated the teams to fortify technical and business relationships.Architected and developed the creative insights and recommendation engine platform for more than 700 enterprise clients including the world's most valuable brands like Amazon and Hilton using 10+M campaign images and 100+ billion email events and marketing campaign data. Mentored and motivated the team to extend their technical/soft skills and helped them map their career progression. Provided them continuous learning and development opportunities including awareness about the company’s core values of empathy, grit and innovation and Diversity, Equity and Inclusion in the workplace.

Sarasota, Fl, Us

Founded an industry-leading predictive marketing solutions company for real estate and related industries which achieved $25M in ARR. Instrumental in SmartZip’s acquisition by Constellation Software (TX:CSU) in 2019. Developed a first of its kind analytics and ML-driven SaaS platform for targeted real estate marketing. The platform was used by thousands of realtors, brokerages and mortgage lenders, utilizing Smart CRM and automated marketing with Google AdWords/Facebook Ads and by direct mail.Created a predictive real estate transaction propensity model named SmartTargeting for 120M+ homes in the US. Created a robust patent-pending set of real estate data products such as: nationwide hierarchical home price model - home price AVM, rental AVM, Neighborhood Price Index and predictive models for 35,000 neighborhoods and 120M homes used by leading real estate portals, lenders, and brokerages. Built a consistent, high scale, reliable, and maintainable data pipeline to process millions of housing data ingested daily into S3/RDS/Redshift using EMR.Created a Direct Mail platform to automate the approval, pdf proof validation, and delivery of physical marketing materials using server-less / Lambda implementations.Architected and developed scalable solutions for a real-time AI-based recommendation notification framework, self-serve geo-spatial tool for real estate insights, reporting dashboards and internal applications used for configuration pricing and quoting. Built a team of over 100 engineers from the ground up, including software engineers, data scientists, and product managers.Helped over 20,000+ clients and 100+ enterprise clients in growing their real estate business while saving their marketing spend. Managed the acquisition and integration of the reputation management company Reach150 with SmartZip.Raised Venture Financing from leading VCs including Intel Capital, Claremont Creek, and Javelin Ventures.

Productized a computer vision-based enterprise security and industrial safety software. Created object recognition and tracking modules for improving the safety of logistics operations by automatically tracking various processes regularly.

Us

Developed ML based recommendation methodologies that rank available travel options by price, convenience and annual contracts. Awarded US Patent ‘Adaptive travel purchasing optimization system’ #US6477520.

Dublin, Ie

Co-founded and directed research and development of corporate travel management software. Developed complex methodologies that rank available travel options by price and convenience. Major participant in the design of a new travel management solution called ‘Cognizer’ that was integrated with existing travel technologies. Awarded US Patent ‘Adaptive travel purchasing optimization system’ #US6477520. Yatra was acquired by Datalex in 2001.

Oversaw research on document analysis and recognition systems for office automation. Managed a German government funded document analysis project called ‘READ’ and published a pair of books along with widely cited research papers.

Bachelor Of Engineering, Electrical Engineering

Ph.D., Computer Science

Masters, Electrical Engineering