Available to hire
Hi, I’m Sahil Padyal, a software engineer specializing in backend systems, AI integration, and cloud-native architectures. I enjoy building scalable services, optimizing performance, and collaborating across teams to deliver impactful products.
In my recent roles, I’ve led efforts in fine-tuning LLMs, building RAG pipelines, and deploying robust CI/CD pipelines on AWS and GCP. I’m passionate about leveraging AI to solve real-world problems and continuously learning new tools and methodologies.
Skills
Experience Level
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Language
Javanese
Advanced
English
Fluent
Work Experience
Software Engineer AI/ML at Humanitarian AI, United States
January 1, 2025 - PresentBuilt a no-code platform to fine-tune open-source large language models from HuggingFace. Collaborated with ML engineers to create pipelines for fine-tuning smaller LLMs, such as Llama3 and Deep Seek, using Unsloth PEFT techniques like LoRA and QLoRA. Implemented a Retrieval-Augmented Generation (RAG) module synthesizing a database of 5,000 entries leveraging Gretel. Designed a secure multi-click deployment workflow using MLflow deployments on AWS SageMaker. Integrated Python FastAPI/Pydantic with AWS Lambda, API Gateway, Cognito, and CDK for user access and metadata management. Key skills used include Transformers, PyTorch, Groq, Python, Unsloth, MLflow, and HuggingFace.
Software Engineer AI at Matrix Rental Solutions, United States
December 1, 2023 - August 18, 2025Engineered a context-aware AI agent-powered conversational chatbot using LangChain and PaLM-2 LLM on GCP Vertex AI. Integrated advanced reasoning, memory, and tool orchestration to enable dynamic, context-aware interactions, boosting user engagement by 40%. Improved retrieval accuracy by 60% through designing and implementing a robust RAG pipeline and Vector DB. Increased model accuracy by 46% by fine-tuning with 10,000 prompts to effectively handle vague and ambiguous queries. Developed an OCR+LLM pipeline to extract income details from payslips, enabling instant and accurate rental eligibility decisions based on income-to-rent ratios, reducing manual review time by 90% and improving approval consistency across diverse document formats. Achieved 40% faster deployment by enhancing CI/CD workflows, leveraging Docker containerization, and implementing pipelines with Cloud Build. Built a user-friendly chatbot UI integrated with the existing platform using React/TypeScript and a scalable b
Software Engineer Backend at Lambda BlueTickVR, India
June 1, 2022 - August 18, 2025Spearheaded 0 to 1 development of a 3D mesh optimization tool leveraging Quadratic Error Metrics (QEM) to automatically reduce polygon count while preserving geometric detail. Architected a distributed system using Ray and Redis for task queuing and caching deployed in Kubernetes (EKS) to handle 100k daily active users, leveraging multi-threaded Python services. Enhanced system functionality by 25% by implementing a scalable backend with Python FastAPI and optimizing read/write operations across RDS PostgreSQL and MongoDB. Optimized deployment workflows by building a CI/CD pipeline with AWS CodeBuild and ArgoCD, utilizing CloudFormation and CDK for consistent, automated infrastructure provisioning. Reduced operational costs by 15% through strategic architecture transition from EC2 to Fargate. Ensured data integrity and security with a security layer using AWS Lambda for file downloads via CDN. Key skills used include FastAPI, AWS CodeBuild, Fargate, S3, AWS Lambda, Kubernetes, ArgoCD,
Software Engineer
October 1, 2019 - August 18, 2025Developed a scalable 3D SaaS inventory tool for real estate builders using Java Spring Boot (MVC pattern), ReactJS, and WebGL, generating $500,000 in initial revenue while enhancing 3D visualization and user experience. Optimized performance by 40% through advanced data structures, algorithms, and object-oriented programming, ensuring high efficiency, early project completion, and robust reliability with JUnit testing.
Software Engineer AI/ML at Humanitarians AI, United States
January 1, 2025 - PresentBuilt a no-code platform to fine-tune open-source LLMs from Hugging Face. Collaborated with ML engineers to fine-tune smaller LLMs like LLaMA 3 and Deep Seek using PEFT techniques (LoRA and QLoRA). Implemented a RAG module synthesizing a 5,000 entry database leveraging Gretel. Designed a secure multi-click deployment workflow using MLflow deployments to AWS SageMaker. Integrated Python FastAPI/Pydantic with AWS Lambda, API Gateway, Cognito, and CDK for user access and metadata management.
Software Engineer AI at Matrix Rental Solutions, United States
December 1, 2023 - August 30, 2025Engineered a context-aware AI agent-powered conversational chatbot using LangChain and PaLM-2 LLM on GCP Vertex AI. Integrated advanced reasoning, memory, and tool orchestration to enable dynamic, context-aware interactions, boosting user engagement by 40%. Improved retrieval accuracy by 60% through a robust RAG pipeline and Vector DB implementation. Increased model accuracy by 46% by fine-tuning with 10,000 prompts to handle vague and ambiguous queries. Developed an OCR+LLM pipeline to extract income details from payslips, enabling instant rental eligibility decisions and reducing manual review time by 90%. Achieved 40% faster deployment by enhancing CI/CD workflows, leveraging Docker for containerization, and implementing pipelines with Cloud Build. Built a user-friendly Chatbot UI integrated with existing platforms using React/TypeScript and a scalable backend built on FastAPI.
Software Engineer Backend at Lambda BlueTickVR, India
June 1, 2022 - August 30, 2025Spearheaded 0 to 1 development of a 3D mesh optimization tool leveraging Quadratic Error Metrics (QEM) to automatically reduce polygon count while preserving geometric detail. Architected a distributed system using Ray and Redis for task queuing and caching deployed in Kubernetes (EKS) to handle 100k daily active users leveraging multithreaded Python services. Enhanced system functionality by 25% by implementing a scalable backend with Python FastAPI and optimizing read/write operations across RDS PostgreSQL and MongoDB. Optimized deployment workflows by building CI/CD pipelines with AWS CodeBuild and Argo CD, utilizing CloudFormation and CDK for consistent automated infrastructure provisioning. Reduced operational costs by 15% through strategic architecture transition from EC2 to Fargate. Ensured data integrity and security with a security layer using AWS Lambda for file downloads via CDN.
Software Engineer
October 1, 2019 - August 30, 2025Developed a scalable 3D SaaS inventory tool for real estate builders using Java Spring Boot (MVC pattern), ReactJS, and WebGL, generating $500,000 in initial revenue while enhancing 3D visualization and user experience. Optimized performance by 40% through advanced data structures, algorithms, and OOP, ensuring high efficiency, early project completion, and robust reliability with JUnit testing.
Software Engineer - AI at Humanitarians AI
January 1, 2025 - PresentContributing to open-source Stellis Labs project to build custom fine-tuned domain-specific LLM models for reasoning and generation tasks. Developed a modular pipeline to fine-tune smaller LLMs (Llama3-1B, DeepSeek-1.5B) for high-accuracy domain-specific reasoning. Designed a secure two-click deployment workflow enabling users to launch fine-tuned models to their own AWS SageMaker instances directly from Hugging Face Hub using CloudFormation, OIDC-based role assumption, and SageMaker SDK, ensuring full customer ownership of artifacts, billing, and security boundaries. Implemented secure APIs with Python FastAPI/Pydantic on AWS, integrating Lambda, API Gateway, Cognito, and CDK to support user access and metadata management. Built a RAG module to synthesize a domain-specific database of 5000 entries using open-source LLMs like Llama3-70B deployed via Groq API. Leveraged Hugging Face Transformers, PEFT, Unsloth, and MLflow for efficient, low-latency fine-tuning and experiment tracking. C
Software Engineer Intern (Generative AI) at Matrix Rental Solutions
December 1, 2023 - September 29, 2025Engineered a high-performance, scalable AI agent-powered chatbot using LangChain and PaLM-2 LLM on GCP Vertex AI, enabling dynamic memory, tool usage, and reasoning to drive 40% higher user engagement. Improved context comprehension by 60% through a custom preprocessing pipeline for prompt templating, token filtering, and conversational history embedding. Increased model performance by fine-tuning on 10,000+ annotated prompts, improving handling of vague/ambiguous queries by 46%. Evaluated retrieval accuracy using Ragas, improving grounding and answer faithfulness by 35% through retriever tuning and prompt optimization. Integrated LangSmith to trace LLM behavior, monitor chain performance, and diagnose tool failures, reducing hallucination rates by 30%. Added moderation layers using Google Perspective API and custom filters to flag toxic queries, reducing offensive outputs by 90%. Reduced data retrieval time by 20% via optimized microservice communication using RESTful APIs (FastAPI/Fl
Software Engineer Backend at Convrse.ai
June 1, 2022 - September 29, 2025Spearheaded 0-to-1 development of a 3D mesh optimization tool using Quadric Error Metrics (QEM) to intelligently reduce polygon count while maintaining visual fidelity. Architected a distributed system with Ray and Redis for task scheduling and caching, deployed on Kubernetes (EKS) to support 100K+ daily active users. Optimized relational and non-relational data performance via RDS PostgreSQL and MongoDB, ensuring efficient schema design and transactional consistency. Developed CI/CD pipelines using AWS CodeBuild, Argo CD, and GitHub Actions, achieving faster deployments and reducing manual deployment errors. Migrated from EC2 to AWS Fargate, cutting infrastructure costs and improving elasticity for compute-intensive workloads. Implemented a secure file distribution layer using AWS Lambda and S3 with signed URLs/CDN integration. Built robust Fast APIs for client-server communication, supporting flexible queries while reducing frontend over-fetching and under-fetching problems. Containe
Software Engineer at Hexaware Technologies
October 1, 2019 - September 29, 2025Developed a scalable 3D SaaS inventory platform for real estate builders using Java Spring Boot, ReactJS, and WebGL, driving initial revenue and significantly improving 3D visualization and customer engagement. Ensured application reliability by developing JUnit test suites, improving backend robustness and reducing post-deployment bugs. Designed and implemented RESTful APIs to facilitate frontend-backend interaction, enabling real-time inventory updates and enhanced system responsiveness. Utilized MySQL for relational database management, ensuring data consistency and scalable storage for inventory records and 3D asset metadata.
Software Engineer - AI at Humanitarians AI
January 1, 2025 - PresentContributes to open-source Stellis Labs project to build custom fine-tuned domain-specific LLMs for reasoning and generation tasks. Designs modular pipelines to fine-tune smaller LLMs (Llama3-1B and DeepSeek-1.5B) for high-accuracy domain reasoning. Implemented a secure, two-click deployment workflow enabling users to deploy fine-tuned models to AWS SageMaker directly from Hugging Face Hub using CloudFormation, OIDC-based role assumption, and SageMaker SDK. Built secure APIs with Python FastAPI/Pydantic on AWS (Lambda, API Gateway, Cognito, CDK) for access and metadata management. Implemented a Retrieval-Augmented Generation module over a 5000-entry domain database using Llama3-70B via Groq API. Employed Hugging Face Transformers, PEFT, Unsloth, MLflow for efficient fine-tuning and experiment tracking. Created RESTful APIs with Flask and stored metadata, interactions, and embeddings in SQL databases. Conducted prompt engineering experiments and evaluated outputs with BLEU, ROUGE, and e
Software Engineer Intern (Generative AI) at Matrix Rental Solutions
December 1, 2023 - September 29, 2025Engineered a high-performance, scalable AI agent-powered chatbot using LangChain and PaLM-2 LLM on GCP Vertex AI, enabling dynamic memory, tool usage, and reasoning to drive 40% higher user engagement. Improved context comprehension by 60% via a custom preprocessing pipeline for prompt templating, token filtering, and conversational history embedding. Fine-tuned on 10,000+ annotated prompts, improving handling of vague/ambiguous queries by 46%. Evaluated retrieval accuracy with Ragas, improving grounding and answer faithfulness by 35% through retriever tuning and prompt optimization. Integrated LangSmith to trace LLM behavior, monitor chain performance, and diagnose tool failures, reducing hallucination rates by 30%. Added moderation layers using Google Perspective API, reducing offensive outputs by 90%. Reduced data retrieval time by 20% via optimized microservice communication using RESTful APIs (FastAPI/Flask) and GraphQL. Enhanced CI/CD by 40% through automated deployment pipelines
Software Engineer Backend at Convrse.ai
June 1, 2022 - September 29, 2025Led 0-to-1 development of a 3D mesh optimization tool using Quadric Error Metrics (QEM). Architected a distributed system with Ray and Redis on Kubernetes (EKS) to support 100K+ daily active users. Optimized relational and non-relational data performance with PostgreSQL and MongoDB. Developed CI/CD pipelines with AWS CodeBuild, Argo CD, and GitHub Actions; Migrated from EC2 to AWS Fargate for cost savings and elasticity. Implemented a secure file distribution layer using AWS Lambda and S3 with signed URLs and CDN integration. Built robust FastAPI backends for client-server communications and containerized microservices with Docker and Helm charts, orchestrating rollouts on Kubernetes.
Software Engineer at Hexaware Technologies
October 1, 2019 - September 29, 2025Developed a scalable 3D SaaS inventory platform for real estate builders using Java Spring Boot, ReactJS, and WebGL, driving initial revenue and enhancing 3D visualization. Built JUnit test suites to improve backend robustness and implemented RESTful APIs for real-time updates. Managed MySQL database for inventory and 3D asset metadata.
Software Engineer AI/ML at Humanitarians AI
January 1, 2025 - PresentContributing to an open-source project to build domain-specific LLMs; developing fine-tuning pipelines for smaller models; implementing a RAG module to synthesize a 5000-entry database; utilizing Hugging Face tools for modular, GPU-optimized implementations; building a REST API with Python Flask and a metadata SQL store.
Software Engineer Intern (Generative AI) at Matrix Rental Solutions
December 1, 2023 - September 29, 2025Engineered a high-performance AI agent-powered conversational chatbot using LangChain and Palm-2 on GCP Vertex AI; improved context accuracy with robust data processing; achieved faster deployment via CI/CD and Docker-based pipelines; enhanced cross-service communication via REST and GraphQL.
Software Engineer Backend at Convrse.ai
June 1, 2022 - September 29, 2025Spearheaded 0-to-1 development of a 3D mesh optimization tool; designed a distributed task queue and caching layer with Ray and Redis deployed on Kubernetes to support 100k daily active users; rebuilt UI with React/Next.js and TypeScript; optimized data management with PostgreSQL and MongoDB; implemented CI/CD with AWS CodeBuild and Argo CD; migrated from EC2 to Fargate to reduce costs; reinforced security with AWS Lambda-powered file downloads.
Software Engineer at BlueTickVR
October 1, 2019 - September 29, 2025Developed a scalable 3D SaaS inventory tool for real estate builders using Java Spring Boot and React; improved performance with advanced data structures and comprehensive testing; delivered on time with robust reliability.
Education
MS in Information Systems at Northeastern University, Boston
September 1, 2022 - December 1, 2024MS in Information Systems at Northeastern University, Boston
September 1, 2022 - December 1, 2024Master's in Information Systems at Northeastern University
January 11, 2030 - September 29, 2025Master’s in Information Systems at Northeastern University
January 11, 2030 - September 29, 2025Master of Science in Information Systems at Northeastern University, Boston
September 1, 2022 - December 1, 2024Qualifications
Industry Experience
Software & Internet, Real Estate & Construction, Financial Services, Professional Services, Education, Computers & Electronics, Telecommunications, Media & Entertainment
Skills
Experience Level
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Hire a AI Engineer
We have the best ai engineer experts on Twine. Hire a ai engineer in New York today.