AI Engineer

As an AI Engineer at BGO, you will work hands-on across the AI stack, building LLM-powered applications, agentic workflows, RAG pipelines, fine-tuned models, and Human-in-the-Loop (HITL) systems that go into production for both BGO's internal operations and our enterprise clients. You will work closely with senior and principal engineers, product managers, and domain experts to take problems from ambiguous ideas to shipped products in days to weeks, not months.

You should be comfortable writing production code, prototyping against frontier models, wiring AI into real operational workflows, and iterating rapidly based on feedback from real users. You do not need to be the deepest researcher in the room, but you do need strong engineering fundamentals, genuine fluency in how modern agentic systems are built, and a strong bias toward shipping.

The core responsibilities for the job include the following:

Hands-On Engineering:

• Build and ship LLM-powered features, agentic workflows, and AI-native product components end-to-end from prompt design and retrieval through APIs, UI integration, and production deployment.
• Implement agentic patterns where they add real value through tool use, multi-step execution, planning, and human-in-the-loop checkpoints with attention to reliability, safety, and clear failure modes.
• Build and operate RAG systems, vector search, hybrid retrieval, and reranking pipelines over BGO and client data.
• Write production-quality Python (and TypeScript/backend code as needed); build APIs and services that integrate cleanly with CRMs, telephony, ticketing, and workflow tools.

AI Products and Client Workflows:

• Contribute to BGO's AI-native products and platform surfaces: Personal AI Partner, Agentic Workflows (QBRs, drafts, and follow-ups), Real-time Ops AI, and Client-Facing AI.
• Partner with forward-deployed engineers and domain experts to customize and deploy AI solutions inside client environments; take client feedback back into the core platform.
• Build reusable, configurable components so new client deployments move in days to weeks.

Human-in-the-Loop and Evaluation:

• Design HITL workflows, review queues, approval steps, escalation rules, and active-learning feedback that let humans supervise, judge, and handle the hard 5% while agents do the heavy lifting.
• Build evaluation harnesses, offline test sets, and online metrics, so we know whether a model or agent is actually working in production.
• Monitor production AI systems; diagnose hallucinations, latency regressions, tool-use failures, and drift; ship fixes fast.

Collaboration and Craft:

• Work closely with product, design, domain experts, and operations to translate real BGO and client workflows into shipped AI features.
• Participate in code reviews, design reviews, and RFC discussions; give and take feedback generously.
• Contribute to BGO's engineering standards for prompts, evals, observability, and responsible AI.

Requirements:

• Bachelor's degree in computer science, engineering, mathematics, or a related technical field (or equivalent practical experience).
• 3-5 years of professional software engineering experience, with at least 1-2 years shipping AI / ML / LLM features in production.
• Strong Python engineering skills; comfort building APIs, services, and data pipelines (REST/gRPC, common cloud primitives).
• Demonstrated experience building LLM-powered applications with real production constraints: latency, cost, reliability, and security.
• Practical, hands-on experience with the modern AI stack: prompt engineering, RAG, vector databases, agent orchestration frameworks (LangChain, LangGraph, Claude Agent SDK, or equivalents), and basic evaluation techniques.
• Experience deploying AI systems on at least one major cloud provider (AWS, GCP, or Azure) with working knowledge of CI/CD, containers, and observability.
• Track record of shipping end-to-end features: you can point to specific production systems you owned and what impact they had.

Preferred Qualifications:

• Experience at an AI-native startup, a top technology company, or on an applied AI/ML platform team.
• Experience building or contributing to agentic systems, autonomous agents, multi-agent workflows, tool use, planning, and long-horizon memory.
• Experience building Human-in-the-Loop systems for high-stakes use cases (compliance, QA, customer operations, healthcare, financial services).
• Experience with fine-tuning, post-training, or building small language models (SLMs) for specific domains.
• Experience with real-time systems streaming, speech, telephony integrations, or low-latency inference.
• Experience in BPO, contact center, customer operations, collections, insurance, or similar operationally dense domains.
• Open-source contributions, side projects, or public technical writing on agentic AI or applied LLM systems.

Expectation:

• AI-Native Mindset: You treat AI as the default starting point for every task, not a side experiment. You reach for an agent before you reach for a spreadsheet.
• Bias for Shipping: You would rather ship a rough v1 this week and iterate than design the perfect v3 for a quarter.
• Pragmatic Judgment: You know when to use an agent, when to use a simple LLM call, and when to use deterministic code, and you can explain the tradeoffs.
• Craft: You care about code quality, evaluations, observability, and responsible AI because flaky agents erode trust and trust is the whole game.
• Ownership: You take features from ambiguous ideas to working in production without needing to be told each next step.
• Curiosity: You are genuinely interested in BGO's domains and the real operational work that agents will be doing on behalf of real people.
• Collaboration: You partner well with domain experts, product, and other engineers; you give good feedback and take it even better.