AI/ML • Software • Data Engineering
Mission-focused models, data solutions, and custom software.
DEOLIVER LLC builds AI/ML systems, data pipelines, and production software for defense, government, and commercial customers — work that has to be correct under scrutiny and reliable after it ships.
Core Capabilities
What we deliver.
-
AI & Machine Learning Development
Custom statistical, machine-learning, and deep-learning models for classification, forecasting, detection, localization, decision support, and NLP.
-
Computer Vision & Sensor Analytics
Solutions for image, sonar, sensor, and time-series data, including feature engineering, detection, tracking, association, and localization.
-
Data Engineering & MLOps
End-to-end data and ML pipelines, production model integration, scalable data processing, and cloud deployment using Snowflake, Databricks, AWS, Azure, and Google Cloud.
-
Custom Software Development
Production applications, APIs, backend services, data applications, and AI-enabled software built to customer specifications.
Relevant Experience / Key Personnel
Prior key-person work across defense, research, telecom, and commercial AI.
-
Defense Logistics Agency support via Sybal
As an AI/ML subcontractor to Sybal on DLA work, increased document sentence-classification accuracy from 54% to 95%, improved cross-document sentence search performance by 10x using vector embeddings and FAISS, and integrated ML outputs into AWS SageMaker for automated compliance verification.
-
University of Texas Applied Research Laboratories
As an ARL employee, developed ML algorithms using sonar imagery and time-series data; improved underwater ordnance localization accuracy by 300% (4x) through feature engineering and Euclidean-distance clustering adaptations; developed feature-embedding association methods to improve detection and tracking of submersed objects.
-
Telecommunications ML / Root Cause Analysis
Develops end-to-end ML software, models, and Snowflake pipelines supporting root-cause analysis across RF, CORE, and TRANSPORT network domains.
-
Commercial AI Applications
Built RAG pipelines and scalable AI applications using Redis, MongoDB, FastAPI, LangChain, LangGraph, and PandasAI; fine-tuned and deployed LLMs across Azure, Google Cloud, and local environments.
Relevant experience above includes prior key-person experience and is not represented as past performance of DEOLIVER LLC unless specifically stated.
Differentiators
Why this work is distinct.
- Direct defense-domain experience supporting DLA work and developing underwater sensing/localization algorithms at a DoD-focused research laboratory.
- Quantified ML performance improvements: 300% (4x) localization accuracy improvement, 54%→95% NLP classification accuracy, and 10x document-search performance.
- Full-stack ML capability spanning data engineering, model development, APIs, cloud deployment, and production integration.
- Experience across sonar, imagery, time-series, telecom network, financial, and large-scale document datasets.
Technical Stack
Languages, platforms, and methods in active use.
- Python
- SQL
- C++
- PyTorch
- TensorFlow
- NLP
- RAG
- LLMs
- Time-Series Modeling
- Computer Vision / Sonar
- Snowflake
- Databricks
- AWS
- Azure
- Google Cloud
- FastAPI
- Docker
- Linux
- Spark
- FAISS
- CUDA
Credentials
-
SnowPro Core Certified
Snowflake platform certification.
-
U.S. DoD Secret clearance
Previously held.
-
M.S., Computational Finance & Risk Management
University of Washington.
Company Data
Legal and contracting identifiers.
- Legal name
- DEOLIVER LLC
- Entity type
- Texas Limited Liability Company
- UEI
- RFHUNHSS4XP8
- CAGE
- 24Z19
- Primary NAICS
- 541511 — Custom Computer Programming Services
- Additional NAICS
- 518210 — Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services
Contact
Daniel Oliver, principal.