About Data Technology and Services
Based in Denver, Colorado, Data Technology and Services is an AI-native engineering firm. We design, build and run production AI systems, including agentic applications, generative AI, computer vision and machine learning, along with the data platforms they depend on. Our clients range from Fortune 500 enterprises to early-stage startups in scientific research, manufacturing and fabrication, financial services, healthcare, medical devices, logistics and consumer markets. AI is where every engagement starts, and we choose the surrounding technology to fit the problem, not the other way around.
What We Build
Custom AI, built from the ground up. We design and build AI applications end to end, combining computer vision, generative AI, machine learning and deep learning to fit the problem instead of forcing it into a packaged product. Our solutions cover demand forecasting, classification, clustering, design QA/QC and supply chain analytics in scientific, manufacturing and fabrication, financial, healthcare and medical device settings, among others. A recent example is a suite of 21 AI applications that automates design review and QA/QC for engineering documents. It combines computer vision, Gemini and Claude and agent workflows connected through Model Context Protocol (MCP) servers. In healthcare, we use foundation and open-source LLMs to turn ambient physician/patient encounters into standardized FHIR/HL7 records, automating more of the administrative workflow.
Machine learning and predictive analytics. We build demand-forecasting pipelines for aerospace supply chains using random forest and regression models, and predictive models that use deep learning and non-linear regression to improve operational efficiency and reduce downtime. For a geopolitical war-game simulation, we used BERT sentence embeddings over the global GDELT news dataset to link related events and themes and model how world events affect supply-chain resilience.
Knowledge graphs, Graph Data Science and GraphRAG. Where relationships matter, we ground AI in a knowledge graph so answers can be traced back to their sources. One example is a Neo4j graph with 80M+ genealogy relationships and 250M+ sampling records that gives a multi-line manufacturer sub-second product traceability. For Entegris, we integrated Graph Data Science into the fabrication process to identify at-risk batches and support impact analysis when fabrication faults occur. Another graph analyzes roughly 70 million research articles for a global publisher.
Real-time data and platform engineering. AI is only as good as the data feeding it. We build event streaming on Apache Kafka, AWS MSK and Flink with change data capture, and we modernize the systems underneath it all. One modernization moved a 30-year-old cost accounting platform with 1M+ lines of code and 1B+ financial records to the cloud and pulled out 200+ KPIs for analytics.
AI Since Before It Was a Buzzword
Our machine learning work goes back more than a decade, to when NLP, embeddings and recommendation engines were built largely by hand:
- WAVi (2007–2014): as CTO of this brain-measurement medical device company, led the EEG signal processing, high-dimensional clustering of time-series data and a SaaS platform designed for thousands of clinics, including SNOMED medical taxonomy for patient records.
- TravelShark (2012–2014): a recommendation engine that used NLP sentiment extraction and unsupervised topic modeling on hotel and restaurant reviews to build an embedding space for “show similar” recommendations.
- dSide (2014–2015): a multi-tenant product recommendation engine that placed products in a vector space to measure user preference, years before vector databases went mainstream.
- Live event analytics: near real-time NLP classification of streaming tweets and texts for consumer brand events, including Lara Bar and Nike marathon activations.
Those projects sit on two decades of earlier large-scale data engineering, from parallel database systems for FedEx and Equifax to enterprise data warehousing for PepsiCo.
Why Choose Us?
Data Technology and Services combines deep industry knowledge with hands-on AI engineering. You work directly with a senior architect, from the first whiteboard session through production deployment.
- 35+ years of hands-on architecture, from enterprise systems for PepsiCo, Kaiser Permanente, FedEx and Charles Schwab to today's AI platforms.
- AI that ships: production agentic and computer-vision applications in daily use, not proofs of concept.
- Machine learning since 2010: NLP, topic models, embedding-based recommendations and biomedical signal analysis, all deployed well before the current wave of generative AI.
- A broad, vendor-neutral stack: Claude, Gemini and open-source LLMs; TensorFlow and Python ML; Neo4j; Kafka, Flink and Spark; BigQuery, AWS and Google Cloud.
- Graph depth when you need it: Neo4j Certified Professional with Graph Data Science certification, and a senior consultant on Neo4j engagements since 2021.
Our Team
Owen Robertson
Principal at DTS
Hands-on architect and solution executive who turns complex enterprise data into knowledge graphs and AI agents that ship. Over more than 35 years, Owen has been a founding member of Tanning Technology, an early big-data consultancy that went public on NASDAQ; CTO of a medical-device startup; and led solution architecture and graph consulting on Neo4j engagements. He has led data architecture and performance work for PepsiCo, Kaiser Permanente, FedEx and Charles Schwab, and most recently built a 21-application AI suite that automates design QA/QC. Neo4j Certified Professional; named inventor on a patient data management patent.
Advisory Board
Alex Robertson
Advisory Board · Semiconductor & Photonics
Physical chemist and semiconductor laser innovator with more than three decades in optoelectronics. As a senior design engineer in the fiber-optics division of a leading global semiconductor company, Alex develops the high-speed lasers behind today’s data-center links, including electroabsorption-modulated lasers for 200G- and 400G-per-lane optics. His research, from the surface chemistry of epitaxial crystal growth to wafer-level laser testing, spans more than 40 publications, 600+ citations and two patents. PhD in physical chemistry, Princeton University; undergraduate degree, Colgate University.
Ambarish Jash
Advisory Board · Applied AI & ML
Accomplished AI and machine learning researcher and engineer with more than 10 years at Google Research, where his work spanned large language models, personalization, embeddings, recommendation systems and production-scale machine learning. His expertise bridges advanced AI research and practical system development, with a particular focus on translating emerging machine learning techniques into scalable, real-world solutions. As an advisor, Ambarish provides guidance on AI/ML architecture, model development and the application of advanced machine learning to complex engineering and industrial problems.