Reza Jelveh

Working hands-on with pre-seed, seed and Series A companies including one Fortune 500 automotive leader, to unstuck struggling Startups quickly turn around their DevOps, MLOps and engineering teams when necessary, as hands-on CTO as a Service and DevSecOps expert. Clients included AI/ML/NLP startups as well as the autonomous vehicle division at Caterpillar.

For Medical startup, built full-stack LLM RAG application for Medical Contract Processing

Directly managed teams of up to 10, and indirectly orgs of 200

At a Fortune 100, directly managed the senior management team for their self-driving AI platform, impacting 100 engineers

At a real-estate-tech seed stage startup, operated as their CTO, restructured the engineering team, doubled team size with new hires, restructured the codebase, introduced a new frontend, and helped release v1 of the product after 2-year delay

At a legal-tech seed stage startup, provided a 100x code reduction and 20x CI performance by rewriting the entire CI pipeline

At a payment-AI seed stage startup, worked with the CTO to rewrite and migrate the entire infrastructure to a fully automated whitelabel product

One of the most ambitious mining projects in the world in partnership with the second-largest mining corporation of the world. This project was aimed to massively increase safety and productivity.

Built and led team of 8 people with seismologist, ML engineers, frontend and backend devs across 5 countries

Architected a scalable containerized solution for a seismic analytics platform for real-time sensor analysis

Introduced a Django-based REST API for the unified access of processed data. Built a real-time web integration in Django

Enhanced a major performance and architectural review and rewrites of all components involved. Rewrote and architected various pieces of Python code into a releasable versionable library

Reverse-engineered a legacy database and storage systems for integration in a Kafka streaming re-architecture, significantly reducing the latency in the data access. I built a Kafka-streaming solution for ingesting legacy data in Kotlin

Developed a Go connector to extract legacy data for stream processing in Kafka

Implemented near real-time processing of seismic events. Also, I matched requirements with potential streaming solutions Flink, Pachyderm, Argo, Airflow, NiFi, Spark, and Flink to provide the best solution to the client

Rearchitected the application to reduce the amount of data passed around in messages and removed and replaced Kafka dependency for normal messaging to Kafka streams processing where it was necessary

Containerized the various components written by scientists into services that can be built and ran equally on all environments

Profiled and rewrote major parts of the application, turning them into a Python library, and added Python packaging and testing infrastructure with corresponding build pipelines

Created Gitlab CI pipelines and migrated everyone to work in a reproducible manner

Designed and implemented a scalable on-premise Kubernetes-based cluster with deployment in a copper mine. It mirrored the deployment on the cloud for developers without the need for access to the internal corporate processes