Syed Akrama

Senior AI/ML Engineer

BigStep technologies • Fulltime

Sep 2025 - Present • 11 mos

Working as a Senior AI/ML Engineer building enterprise-scale AI platforms, LLM-powered applications, and production-grade MLOps systems deployed on cloud infrastructure. My work focuses on designing reliable AI systems that combine machine learning models, large language models, and data pipelines to automate complex business workflows. • Designed and deployed LLM-driven AI platforms integrating retrieval-augmented generation (RAG), vector search infrastructure, and agent-based workflows to enable intelligent document processing and enterprise knowledge automation. • Architected production-grade document intelligence systems capable of extracting structured data from large volumes of unstructured documents including financial statements, legal contracts, invoices, and scanned PDFs. • Built entity resolution and semantic matching systems using hybrid vector embeddings and similarity models to unify fragmented enterprise datasets across multiple systems such as CRM and data warehouses. • Implemented agentic AI workflows capable of multi-step reasoning, tool usage, and API orchestration to automate operational processes across enterprise data platforms. • Designed scalable AI backend services using Python and FastAPI, enabling machine learning models and LLM capabilities to be exposed through robust REST APIs. • Developed MLOps pipelines using AWS SageMaker, implementing automated training, evaluation, deployment, and monitoring workflows for machine learning models in production environments. • Implemented data quality monitoring, model validation, and drift detection systems to ensure reliability, reproducibility, and governance of AI pipelines operating on large datasets. • Designed data validation and pipeline safeguard frameworks to prevent silent data corruption and ensure high integrity for downstream machine learning models. • Built scalable data processing pipelines using distributed frameworks and cloud storage systems, enabling high-performance data ingesti

Lead Software Engineer - AL/ML Branch

UsefulBI Corporation • Fulltime

Jun 2023 - Sep 2025 • 2 yrs 3 mos

Designed and deployed production-grade AI systems focused on LLM applications, agentic workflows, and large-scale data platforms used in enterprise environments. • Built agentic AI systems and multi-step retrieval-augmented generation (RAG) pipelines using large language models to automate document understanding and enterprise knowledge retrieval. • Developed LLM-powered document intelligence platforms capable of extracting structured data from complex financial and legal documents, including scanned PDFs and handwritten annotations. • Implemented fine-tuning and prompt optimization workflows for domain-specific models using techniques such as parameter-efficient fine-tuning (PEFT / LoRA) and retrieval grounding to improve extraction accuracy and response reliability. • Architected vector search and embedding pipelines to enable semantic document search and contextual reasoning across large enterprise datasets. • Designed multi-agent orchestration systems that integrate external APIs, enterprise tools, and reasoning workflows to automate complex operational tasks. • Built scalable backend AI services using Python, FastAPI, and asynchronous processing frameworks to expose LLM capabilities through production APIs. • Developed MLOps pipelines on AWS and Databricks, including automated training pipelines, evaluation frameworks, and model monitoring. • Implemented data quality validation, drift detection, and automated pipeline safeguards to ensure reliability of AI systems operating on large production datasets. • Optimized distributed data processing pipelines using PySpark and Databricks, significantly reducing processing latency for large-scale analytics workloads. • Collaborated with cross-functional teams to translate complex business requirements into scalable AI architectures and deploy production-ready machine learning solutions.

Deployed Machine Learning Engineer

MAQ software • Fulltime

Mar 2022 - Jun 2023 • 1 yr 3 mos

Worked as an Applied Machine Learning Engineer delivering production-grade AI and data science solutions for enterprise use cases, with the primary client being Microsoft. My work focused on building scalable machine learning models, data pipelines, and analytics systems operating on large datasets to support business decision-making and operational automation. • Designed and deployed machine learning models for Microsoft enterprise platforms, working with multi-million record datasets to support internal analytics and decision intelligence workflows. • Built recommendation systems using neural network models to suggest relevant enterprise services and resources, improving engagement and personalization across Microsoft internal tools. • Developed NLP pipelines using BERT-based architectures for sentiment analysis and text classification aligned with Microsoft business taxonomies, achieving high classification accuracy across large text datasets. • Implemented predictive modeling and optimization algorithms for resource allocation and budgeting workflows, improving operational efficiency and reducing unnecessary expenditure. • Developed computer vision models using YOLO-based architectures for automated detection tasks within internal automation workflows. • Built scalable data pipelines and ETL workflows to process large volumes of structured and unstructured data using Python, SQL, and distributed data processing frameworks. • Optimized data processing pipelines and analytical workflows, reducing processing latency and enabling faster insights generation for business teams. • Developed interactive dashboards and analytics reports using Power BI and SQL, enabling leadership teams to monitor KPIs, track performance metrics, and evaluate machine learning model impact. • Collaborated closely with Microsoft stakeholders and cross-functional teams to translate ambiguous business requirements into production-ready AI and data solutions. • Delivered deployed machine lear