Hi, I'm Emad.
I build intelligent systems.

Designed and built an on-demand model-retraining pipeline on SageMaker Pipelines, turning researchers' feature-generation, training, and evaluation code into reusable DAG steps. Enabled parallel experiments tracked in MLflow, cutting experiment turnaround by an estimated ~40%. Built proof-of-concept Snowflake processing and training pipelines and assessed the gaps for migrating AWS research workflows to Snowflake.

Contributed to CLEAR Investigate, Thomson Reuters' commercially launched agentic AI product for investigations. Built a central configuration layer in PydanticAI for agent system prompts, tool definitions, and agent/subagent wiring, so the team could reconfigure investigation agents from one place instead of editing definitions across many files. Migrated the LLM-as-judge evaluation suite into a reusable repository prepared for CI/CD.

Built the GenAI summary panel for CLEAR Business entity reports, live in production processing hundreds of reports daily. Designed selective XML extraction feeding targeted per-section LLM calls (business overview, key personnel, adverse media, liens and lawsuits), plus a Bing Search + LLM verification step that filters false matches between same-named entities, with each claim linked to its source in the report. Refined prompts and extraction logic through SME annotation rounds against product-manager exemplars.

Built the research data infrastructure for CLEAR's entity resolution system (~800M entities, billions of documents, 700 idents/second). Designed a PII-isolated AWS research environment with SageMaker Studio, CloudFormation, and per-researcher profiles; defined a versioned schema reconciling two incompatible annotation sources, with Spark and Python pipelines to merge records and surface label conflicts; and automated the manual workflow into scheduled, repeatable SageMaker Pipelines.

Shipped a sentence-embedding query-intent classifier for Checkpoint, Thomson Reuters' tax research platform, separating broad conceptual queries from specific tax lookups to promote higher-level documents — improving access to conceptual documents by 95% within a ~100ms latency budget. Identified the retrieval gap from SME and customer feedback.

Fine-tuned T5 to generate questions from documents, cutting manual data-curation effort by 40%. Experimented with input representations and evaluated generated text with BLEURT in a research collaboration with university partners.

Trained domain-specific Word2Vec embeddings and LSTM-based NLU models, with interpretability testing to make model decisions transparent.

Built a conversational question-answering system using syntactic and semantic document analysis and automatic ontology generation, in a research collaboration with an industry partner.