Erias Ventures
Enterprise - Senior AI/ML Engineer - GEOINT, GNN, Computer Vision
Erias Ventures, Dayton, Ohio, United States, 45444
Overview
Enterprise - Senior AI/ML Engineer - GEOINT, GNN, Computer Vision at Erias Ventures. We are seeking engineers who wish to grow their careers and want to become part of a technically strong and growth-oriented company focused on bringing innovative solutions to the difficult mission problems facing our customers. Responsibilities
Design and fine-tune vision models for a government GEOINT Scene Prediction prototype. Apply graph-neural network context reasoning and build searchable embeddings with vector storage/retrieval. Architect training pipelines and contribute to low-SWaP prototypes; deliver auditable memory services for satellite/GEOINT data. Support the full lifecycle from R&D through deployable prototypes and tech transfer. Clearance
A current Top-Secret/SCI security clearance is required. Candidates cannot be sponsored or nominated for a government security clearance under this position. Experience & Qualifications
Bachelor’s degree from an accredited college or university in CS/EE/Math or related field; Master’s may substitute for 2 years of experience. Five (5) years total experience in applied ML/computer vision or closely related field. Required Skills
Computer Vision & VLMs: Hands-on experience with YOLO (YOLOv8-OBB or rotated-box detectors) and at least one of DETR, ViTDet, or VLMs for overhead imagery; fine-tuning models on satellite/GEOINT datasets; post-processing for orientation/heading estimation and spatial relationships. Graph Neural Networks (GNNs): Practical understanding of message-passing and building graphs from detection outputs to improve classification, association, and false-alarm reduction. Embeddings & Vector Stores: Experience generating CLIP/VLM embeddings, implementing k-NN search, and configuring FAISS/Milvus for vector storage/retrieval; ability to design metadata schemas, similarity metrics, and analyst-friendly queries. Data Management: Proficiency with SQL (PostgreSQL/Redshift) and NoSQL (MariaDB, MongoDB); experience modeling imagery, objects, actions, and provenance data. Training Pipelines: Skilled in building supervised, semi-supervised, and self-supervised training loops; familiar with parameter-efficient fine-tuning (e.g., QLoRA) for compact models. Prototype & Low-SWaP: Ability to deliver edge-native demos on constrained devices (e.g., Raspberry Pi-class) with quantized models, motion filtering, and local vector indexing. Quality & Audit-ability: Experience with spatial-consistency checks, prediction versioning, audit logs, and measurable improvements in precision/recall. Tooling & Delivery: Proficient in Python/PyTorch; containerization; Jupyter for rapid prototyping; CI/CD workflows; and producing clear, transition-ready documentation. Desired Skills
Experience with open overhead-imagery datasets (e.g., RarePlanes, MAR20, xView, DOTA v2) and applying domain adaptation to new sensors. Experience deploying Milvus with dynamic-field schemas and incremental indexing; familiarity with ChromaDB for lightweight development. Prior work developing analyst-facing retrieval UIs or integrating similarity scores into triage tools. Familiarity with governance and commercialization rights vetting for model transfer to classified environments. Strong communication skills and experience working in agile, mission-focused teams. Benefits
Salary range: $170,000 - $205,000 depending on education and years of experience. Wealth Benefits: Above Market Hourly Pay; 11% 401k with immediate vesting; Spot Bonuses for business development; Professional Development Bonuses. Health Benefits: Company-subsidized medical coverage; 100% company-paid vision and dental; 100% company-paid long-term disability, short-term disability, and group life insurance; Monthly wellness reimbursement. Happiness Benefits: Paid time off with flexible schedules and birthday off; Amazon Prime membership and monthly internet reimbursement; Technology and productivity allowance; Morale events; Onboarding and annual swag; Company-paid professional development and training. We are an equal opportunity employer. Qualified applicants will be considered without discrimination based on race, color, religion, gender, gender identity or expression, sexual orientation, national origin, genetics, disability, age, or veteran status. Referrals & Inquiries
We offer referral bonuses of up to $10,000 for each successful hire. Send referrals and inquiries to jobs@eriasventures.com.
#J-18808-Ljbffr
Enterprise - Senior AI/ML Engineer - GEOINT, GNN, Computer Vision at Erias Ventures. We are seeking engineers who wish to grow their careers and want to become part of a technically strong and growth-oriented company focused on bringing innovative solutions to the difficult mission problems facing our customers. Responsibilities
Design and fine-tune vision models for a government GEOINT Scene Prediction prototype. Apply graph-neural network context reasoning and build searchable embeddings with vector storage/retrieval. Architect training pipelines and contribute to low-SWaP prototypes; deliver auditable memory services for satellite/GEOINT data. Support the full lifecycle from R&D through deployable prototypes and tech transfer. Clearance
A current Top-Secret/SCI security clearance is required. Candidates cannot be sponsored or nominated for a government security clearance under this position. Experience & Qualifications
Bachelor’s degree from an accredited college or university in CS/EE/Math or related field; Master’s may substitute for 2 years of experience. Five (5) years total experience in applied ML/computer vision or closely related field. Required Skills
Computer Vision & VLMs: Hands-on experience with YOLO (YOLOv8-OBB or rotated-box detectors) and at least one of DETR, ViTDet, or VLMs for overhead imagery; fine-tuning models on satellite/GEOINT datasets; post-processing for orientation/heading estimation and spatial relationships. Graph Neural Networks (GNNs): Practical understanding of message-passing and building graphs from detection outputs to improve classification, association, and false-alarm reduction. Embeddings & Vector Stores: Experience generating CLIP/VLM embeddings, implementing k-NN search, and configuring FAISS/Milvus for vector storage/retrieval; ability to design metadata schemas, similarity metrics, and analyst-friendly queries. Data Management: Proficiency with SQL (PostgreSQL/Redshift) and NoSQL (MariaDB, MongoDB); experience modeling imagery, objects, actions, and provenance data. Training Pipelines: Skilled in building supervised, semi-supervised, and self-supervised training loops; familiar with parameter-efficient fine-tuning (e.g., QLoRA) for compact models. Prototype & Low-SWaP: Ability to deliver edge-native demos on constrained devices (e.g., Raspberry Pi-class) with quantized models, motion filtering, and local vector indexing. Quality & Audit-ability: Experience with spatial-consistency checks, prediction versioning, audit logs, and measurable improvements in precision/recall. Tooling & Delivery: Proficient in Python/PyTorch; containerization; Jupyter for rapid prototyping; CI/CD workflows; and producing clear, transition-ready documentation. Desired Skills
Experience with open overhead-imagery datasets (e.g., RarePlanes, MAR20, xView, DOTA v2) and applying domain adaptation to new sensors. Experience deploying Milvus with dynamic-field schemas and incremental indexing; familiarity with ChromaDB for lightweight development. Prior work developing analyst-facing retrieval UIs or integrating similarity scores into triage tools. Familiarity with governance and commercialization rights vetting for model transfer to classified environments. Strong communication skills and experience working in agile, mission-focused teams. Benefits
Salary range: $170,000 - $205,000 depending on education and years of experience. Wealth Benefits: Above Market Hourly Pay; 11% 401k with immediate vesting; Spot Bonuses for business development; Professional Development Bonuses. Health Benefits: Company-subsidized medical coverage; 100% company-paid vision and dental; 100% company-paid long-term disability, short-term disability, and group life insurance; Monthly wellness reimbursement. Happiness Benefits: Paid time off with flexible schedules and birthday off; Amazon Prime membership and monthly internet reimbursement; Technology and productivity allowance; Morale events; Onboarding and annual swag; Company-paid professional development and training. We are an equal opportunity employer. Qualified applicants will be considered without discrimination based on race, color, religion, gender, gender identity or expression, sexual orientation, national origin, genetics, disability, age, or veteran status. Referrals & Inquiries
We offer referral bonuses of up to $10,000 for each successful hire. Send referrals and inquiries to jobs@eriasventures.com.
#J-18808-Ljbffr