Generative AI Specialist

Bright Vision Technologies•Sugar Land, TX

1d•$100,000 - $150,000•Remote

About The Position

Bright Vision Technologies is seeking a Generative AI Specialist to define and lead organization-wide strategy, patterns, and tooling for prompt design, agentic workflows, and LLM-based application architectures. The role combines practical expertise with modern LLMs and the development of reusable design patterns, evaluation frameworks, and developer tools that can scale across teams and use cases. The ideal candidate has shipped LLM-powered products to production, understands prompt-design patterns and agent architectures, and can make sound decisions about applying LLM techniques responsibly. This is a 100% remote, full-time, direct W2 position with Bright Vision Technologies, based in the continental United States. The position supports the company’s in-house Statement of Work engagement; Bright Vision Technologies is the employer and there is no third-party client, vendor, or implementation partner. C2C, 1099, and other third-party arrangements are not accepted. New H1B sponsorship is unavailable, but qualified candidates currently on a valid H1B visa may apply for a transfer. A technical coding assessment is mandatory.

Requirements

Bachelor’s or Master’s degree in Computer Science, Computational Linguistics, or a related field.
Six or more years of software engineering experience, with significant time spent on LLM-based applications.
Demonstrated experience shipping LLM-powered products to production.
Deep familiarity with modern LLM APIs and agent frameworks.
Strong understanding of retrieval-augmented generation, embeddings, and vector databases.
Experience designing evaluation pipelines for non-deterministic systems.
Strong Python skills and comfort with modern application frameworks.
Solid grasp of responsible AI principles, including safety and policy considerations.
Excellent written and verbal communication skills.
Track record of mentoring engineers and influencing technical direction.
Candidates must be willing to work directly as a full-time W2 employee and contribute to in-house SOW deliverables.
Candidates must be confident in their technical abilities and hands-on experience and complete a mandatory technical coding assessment.

Nice To Haves

Public writing, talks, or open-source contributions on LLM application development.
Experience with multi-agent architectures and complex tool-use systems.
Familiarity with fine-tuning workflows and when to choose them over prompting.
Exposure to product domains such as customer support, coding assistants, or analytics agents.
Experience integrating LLMs into enterprise software systems with strict compliance requirements.

Responsibilities

Define organization-wide standards, patterns, and reference architectures for LLM-based applications.
Design prompt structures, instruction templates, and retrieval strategies for production use cases.
Architect agentic systems incorporating tool use, planning, memory, and multi-step reasoning.
Lead the design of retrieval-augmented generation pipelines, including chunking, indexing, and reranking strategies.
Develop evaluation frameworks for prompt quality, agent reliability, and end-to-end task success.
Build internal tooling and libraries to accelerate LLM application development across teams.
Establish guardrails, safety filters, and policy enforcement patterns for LLM-powered products.
Collaborate with model engineering teams on prompt-model co-design and fine-tuning opportunities.
Conduct technical reviews of LLM application designs across product teams.
Mentor engineers and applied scientists on prompt engineering and LLM application architecture.
Lead red-teaming exercises and continuously improve robustness against adversarial inputs.
Track latency, cost, and quality trade-offs in LLM application design and recommend optimizations.
Document patterns, anti-patterns, and lessons learned for internal reuse.
Stay current with LLM capabilities, tooling, and research, and translate advances into practical guidance.