Iman Kohyarnejadfard

Lead applied ML and engineering across multiple products, owning work from problem framing through production deployment. - General time-series forecasting framework • Built a domain-agnostic forecasting framework reusable across products and industries. • Used Prophet, Temporal Fusion Transformer, Autoformer, and PatchTST; parallelized in Databricks to forecast millions of series at scale. • Developed LLM-augmented forecasting — incorporating contextual data and generating predictive insights via LLMs. • Designed data pipelines, quality checks, and Databricks dashboards delivering actionable insights to non-technical users. - Agricultural forecasting platform • Applied the forecasting framework to specialty-crop agriculture: crop-yield forecasts, inventory optimization, and multi-source data tracking. • Built watchlist and filtering features that help farmers and produce suppliers make data-driven decisions. - Political intelligence platform • Designed an LLM agent system with multi-provider orchestration (Claude, GPT, Gemini) that autonomously generates reports, summaries, charts, and polls. • Built NLP pipelines for political bias and sentiment classification, mapping discourse to a structured taxonomy. • Engineered generative-AI personalization at scale via resilient batch inference on serverless infrastructure. • Implemented LLM observability and evaluation (Langfuse, PostHog, OpenTelemetry) and a versioned prompt registry. - Enterprise content automation platform • Shipped a production RAG pipeline (LangChain + OpenAI) with document chunking, embedding, retrieval, and reranking. • Built a content benchmarking engine using BERTopic, spaCy, and sentiment/trend analysis. • Orchestrated fault-tolerant ML workflows with Temporal; delivered an API-first backend and customer-facing frontend.

- Conducted research and developed AI-driven system performance analysis tools, focusing on anomaly detection and tracing data analysis to enhance large-scale distributed systems and microservice architectures. - Researched and developed several novel frameworks for performance analysis and anomaly detection, including: * System Performance Anomaly Detection Using Tracing Data Analysis – Leveraged tracing logs and machine learning models to identify performance degradations in distributed systems. * A Framework for Detecting System Performance Anomalies Using Tracing Data Analysis – Designed a scalable pipeline integrating machine learning and statistical methods to detect abnormal system behavior. * Anomaly Detection in Microservice Environments Using Distributed Tracing Data Analysis and NLP – Developed an AI-powered anomaly detection framework leveraging Natural Language Processing (NLP) and tracing data to enhance observability in microservices. - Designed and implemented system performance monitoring tools using LTTng, Jaeger, and Trace Compass, enabling real-time detection and analysis of performance bottlenecks.

Teaching Assistant of "Big Data Mining"

- Developed and deployed transformative software solutions, enhancing operational efficiency and business value across HEDCo’s energy engineering operations. - Designed and implemented AI-driven automation solutions, streamlining engineering workflows and operational processes to improve efficiency. - Integrated data-driven insights to support energy management, asset monitoring, and decision-making for industrial operations. - Collaborated with cross-functional teams to align software solutions with engineering processes, optimizing automation strategies and improving business intelligence.

- Designed and maintained enterprise software applications for industrial automation, streamlining production workflows and data management. - Developed and integrated database-driven solutions to enhance inventory management and production tracking, improving operational efficiency.

Teacher of "Discrete Event System Simulation"

- Developed desktop and web-based applications for educational management systems, automating administrative tasks and enhancing user experience. - Provided training and mentorship on software development best practices, assisting students in improving their programming skills.

My responsibilities include…

Teaching Assistant of "Logic Circuits" and "Formal Languages and Automata Theory"

Thesis title: System performance anomaly detection using tracing data and Machine Learning