Development and Refinement: Lead the development and continuous improvement of Skylark's large language models, ensuring they are optimized for accuracy, efficiency, and scalability.Product Integration: Work closely with product development teams to integrate LLM capabilities into Skylark's product features, enhancing the overall user experience.Performance Optimization: Implement strategies to optimize the performance of LLMs, including fine-tuning, model compression, and efficient deployment techniques.Collaboration: Collaborate with engineering and design teams to ensure seamless integration of LLMs with other components of the product, incorporating customer feedback and market trends.Research and Innovation: Stay up-to-date with the latest advancements in AI and LLMs, applying this knowledge to drive innovation and improve Skylark's AI capabilities.Data Management: Manage and preprocess large datasets for training and fine-tuning LLMs, ensuring data quality and relevance.Customer Demos and Presentations: Support the product and sales teams by conducting technical demos and presentations, showcasing the capabilities of Skylark's AI solutions to potential clients and stakeholders.Market Analysis: Analyze market trends, customer insights, and competitive landscape to inform product strategy and LLM development priorities.

○ Leverage in-depth experience with GPT models to craft, train, and fine-tune solutions forLLM, emphasizing on specialty in finance and M&A applications.○ Utilize ML expertise in diverse applications, encompassing structured enterprise data,predictive analytics, and insight generation.○ Develop and adjust generative AI structures and strategies such as GAN, GPT, andrelated methodologies tailored to proprietary data engines.○ Collaborate with interdepartmental teams to conceive and deploy avant-garde solutionscapitalizing on LLM and GPT structures.○ Guide and mentor a brigade of 2-5 ML professionals, leading projects from inception tofruition.○ Pioneer R&D efforts in the domains of LLM, GPT, and associated arenas.○ Remain abreast of contemporary breakthroughs in ML, AI, and DS.○ Leverage in-depth experience with GPT models to craft, train, and fine-tune solutions forLLM, emphasizing on specialty in finance and M&A applications.○ Utilize ML expertise in diverse applications, encompassing structured enterprise data,predictive analytics, and insight generation.○ Develop and adjust generative AI structures and strategies such as GAN, GPT, andrelated methodologies tailored to proprietary data engines.○ Collaborate with interdepartmental teams to conceive and deploy avant-garde solutionscapitalizing on LLM and GPT structures.○ Guide and mentor a brigade of 2-5 ML professionals, leading projects from inception tofruition.○ Pioneer R&D efforts in the domains of LLM, GPT, and associated arenas.○ Remain abreast of contemporary breakthroughs in ML, AI, and DS.

Interpersonal and communication skills through the conversations with prospective students as a representative of the Northeastern University - CPS Customer service skills through the process of understanding the questions and working to find the best answers I can for the prospective student. This may include doing research or working with the wider team of staff and Ambassadors Marketing, creative and social media skills through the collaboratively working with the marketing department to create social posts for the Northeastern University - CPSWritten communication skills as all the work I did as an Ambassador through The Ambassador Platform involved written communications, some of which were public-facing Organization, planning and time management skills to be able to fit this additional workaround my studies whilst being able to meet all deadlinesTeam-work and professional skills were required to work collaboratively both with staff at the Northeastern University - CPS and other Ambassadors

Applications of Artificial Intelligence, NLP, Computer Vision, Complex Problem Solving, Enterprise Architecture, Healthcare Information Processing, Data Mining, Advanced Analytics, AI Communications/Visualizations, Research Computing, Data Management& Big Data, Data Wrangling with SAS

1. R language Lab instructor for Biostatistics course at Bouve College of Health Sciences, Northeastern University. Teaching R for Statistical Analysis in field of Public Health.2. Instructing Biostatistics concepts with implementation in R language to 50+ students and contribute on institutional boards3. Resolving doubts about statistical concepts – t-test, F-test, p-value, logistic regression, and Monte Carlo Simulations

1. Solving big business problem of intensive manual work in real estate, transforming complex transformer architectures like T5-large, BERT2. Demonstrate deeper understanding of BERT, T5, BART, LLaMA, for customized automated pipeline for highly secured datasets3. Conduct training and optimizations like pruning, PEFT, LoRA, minimizing sparseness, and establishing methods to perfectly fine tune on limited 15 GB GPU4. High performance computing with GPU accelerators are being utilized for production grade pipeline, in PyTorch and Tensorflow5. Customizations in T5 involved more than 270 trillion Floating Point Operations, after training, work is on streaming it for use case6. Knowledge graphs, BlockChain Graph Neural Networks were also studied for more sophisticated approach

1. Automated the manual process of classifying IBD patients from Clinical notes by incorporating NLP methods and algorithms2. Acquired 5 datasets from EHR data provider over 5,000 exams, including ICD codes (for Crohn's Disease and Ulcerative Colitis) with SAS3. Developed NLP algorithms to automate the identification of patients with active IBD, eliminating the need for manual process4. Employed vectorization techniques like BoW, Word2Vec, and TF-IDF to analyze the frequency and associations of words, all these implemented in R5. Customized Bio-BERT, LLM model with modifying MedSpaCy NER element to increase overall model performance by 7%6. Implemented deep learning models like Recurrent Neural Network(RNN) and ensemble models, achieving 80% F-1 score7. Delivered actionable insights to healthcare professionals through Power BI dashboards, data driven presentations and shared production level Python files, by carefully tailoring technical intricacies into non-technical to easily communicate to Management.8. • Customized NER for domain specific data and developed chatbot with ChatGPT API for data extraction from conversations between doctor and patient9. From Data acquisition to data entry insights, explained through SAS

1. Led a high-performing team to create real-time flight tracking software with FastAPI, deployed ML pipeline with Airflow and AWS EKS for ITQ's edge application(built with Go and RUST/C++) using automation tools like Terraform and Jenkins2. Researched and optimized hyper-parameter tuning process and time-series analysis with an overall 78% efficiency increase3. Applied AGILE Scrum in collaboration with UI/UX team, to deploy model in production, with GIT, saving $1.3M+ on AWS4. Consulted Technical and marketing team to develop condition-based monitoring models, led to a 40% increase in user engagement5. Acquisition of data, and management of company’s own database was handled with SAS programming, also performed statistical analysis about relationships between weather data and Flight delays

1. Analyzed Meta’s (client) performance in SuperheroU ad campaigns. Improved ongoing analysis by employing statistical and probabilistic knowledge, and leveraged data analysis languages like SQL, SAS, Python and R for wrangling and storage purpose. 2. Analyzed Meta’s (client) KPIs of SuperheroU ad campaigns, improved ongoing analysis by employing statistical, probabilistic and data cleaning knowledge, and leveraged data analysis languages like SQL, Python and R for wrangling and storage with SAP3. Developed dashboards using Tableau and Power BI to effectively communicate insights to Stakeholders during team meetings about efficient allocation of $110M+ and increased marketing productivity by 34%, Employed XGBoost for boosting the model4. Proposed customer segmentation, enabled personalized marketing strategies, increasing customer satisfaction by 40%

1. Analyzed activity of partners on Indus App Bazaar Developer Portal and come out with efficient ways to strengthen bond between partners and Indus OS, and expanded User base to 120M+.2. Collaborated with Xiaomi and Samsung for Deep Product integration, shared insights to boost total users to 100 M+ with 900 million+ installations with dashboards and reports, technologies used were, SAS, MS Excel, and Power BI3. Drive focus of Indian indigenous application developers towards Samsung App Store and decreased customer churn rate by 5%4. Improved Indus App Bazaar(India's Largest Indigenous App Store) user base by 350K+5. Drive focus of Indian indigenous application developers towards Samsung App Store and decreased customer churn by 5%.6. Managed to make more indigenous allies to Indus OS for example, SYGO Games, Imangi Studios, Lipikaar and 200 other such app developers.

1. Collaborated with E-Commerce and Innovation department for devising new routes for flights over Asian countries to reduce fuel cost and reduce pollution and time of flight to almost 40%.2. Managed various banners and fair assurance on websites. Actively responsible for updating websites of majority of 12 major countries of Europe and Asia.3. Troubleshoot network and production issues and perform root-cause analysis. Worked under Tech team for firewall updation for in-flight WiFi.