Lead architecture and production of complex multi-agent LLM systems. Design evaluation pipelines (Evals), optimize RAG and vector search for scale, manage cost and latency tradeoffs, translate product requirements into agent behaviors, and mentor junior AI engineers.
About Us
At Turgon, we believe there is a massive opportunity to reimagine how IT services are delivered by leveraging a combination of AI agents and humans to build, integrate, and run the enterprise stack. AI agents that understand your existing infrastructure, processes, and data and can rapidly add new business logic, generate new integrations on the fly, and root cause issues instantly. That’s a $1.5T opportunity!
We are a startup based in San Francisco and New Delhi, founded by serial entrepreneurs. We are building out our core team to re-imagine and build an “AI-Native Accenture.” If you are compelled by this vision, let’s chat!
About the Role
We are looking for a Senior AI Engineer to take ownership of our most complex agentic workflows. In this role, you will bridge the gap between "cool demo" and "enterprise-grade reliability."
You won’t just be prompting models; you will be architecting the orchestration layer that governs them. You will be responsible for the end-to-end lifecycle of our agents—from system design and prompt engineering to latency optimization, cost management, and rigorous evaluation frameworks. You will also mentor younger engineers and help define our AI engineering best practices.
Responsibilities
- Lead the architectural design of complex, multi-agent systems using LangGraph/LangChain that can handle long-running, asynchronous enterprise tasks.
- Own the "Evals": Design and build automated evaluation pipelines to measure agent accuracy, hallucination rates, and success metrics before deploying to production.
- Production Hardening: Optimize RAG pipelines for scale—tuning chunking strategies, retrieval algorithms, and vector search parameters for maximum relevance and speed.
- Cost & Latency Engineering: Make critical decisions on when to use massive foundation models (GPT-4) vs. smaller, specialized models to balance performance with unit economics.
- Collaborate with the Product Team to translate ambiguous business requirements ("Fix the ERP data") into concrete agentic behaviors and tool definitions.
- Mentor junior AI engineers, conducting code reviews and helping them level up their understanding of LLM application architecture.
Requirements
- 5+ years of total engineering experience, with significant recent focus on Applied AI / LLMs.
- Expert-level Python skills; you write clean, modular, and testable code.
- Deep production experience with LLM orchestration (LangChain, LangGraph, etc.) – you know where these frameworks break and how to fix them.
- Strong understanding of the modern Data Stack (SQL, Snowflake, Vector DBs); you know that good AI requires good data.
- Experience in MLOps or LLMOps (tracing, monitoring, and versioning prompts/chains).
- A pragmatic mindset: You know when to use a simple heuristic and when to deploy a complex agent.
Nice-to-Have Skills
- Experience fine-tuning open-source models (Llama, Mistral) for specific domain tasks.
- Background in traditional NLP or Machine Learning prior to the Generative AI boom.
- Experience with Kubeflow, Ray, or other distributed ML computing frameworks.
- Previous experience in a Founding Engineer or early-stage startup role.
Education & Certification
- We don’t actually care about qualifications as long as you are a rockstar engineer who wants to keep learning and make a massive impact.
What We Offer
- A values-aligned culture: always learning mindset, collaborative environment, and an entrepreneurial approach to getting things done.
- Competitive, flexible salary based on experience.
- Equity via ESOPs.
- Leadership role shaping AI/LLM services at enterprise scale.
- Collaborative, high-energy startup culture in South Delhi and San Francisco.
- Access to top AI researchers worldwide to drive innovation and continuous learning.
- Flexible work-from-home days balanced with on-site collaboration.
Similar Jobs
Artificial Intelligence • Software • Analytics • Utilities
Build and ship customer-facing AI features for Scout, including agentic LLM workflows, tool calling, RAG pipelines, evaluations, guardrails, and prompt engineering. Develop full-stack services and interfaces using backend languages, TypeScript, React, and mobile technologies. Deploy scalable, secure systems on AWS and Kubernetes, with monitoring, testing, and CI/CD. Own features end-to-end, collaborate with stakeholders, make product decisions, and contribute to reliable AI infrastructure and engineering practices.
Top Skills:
Agentic WorkflowsAnthropicApexAWSCi/CdCrewaiDockerFastapiGoGpu InferenceJavaKubernetesLangchainLangfuseLanggraphLangsmithLlamaindexLlmsModel Context ProtocolMulti-Agent SystemsNode.jsNosql DatabasesOpenaiPrompt EngineeringPythonRagReactReact NativeRelational DatabasesRest ApisSalesforceTool CallingTypescriptVector DatabasesVllm
Artificial Intelligence • Consumer Web • Edtech • Enterprise Web • HR Tech • Social Impact • Generative AI
Build and deploy AI-powered customer solutions across discovery, prototyping, and production. Responsibilities include designing secure multi-tenant and hybrid architectures, integrating identity and access management, encryption, networking, data engineering, CI/CD, observability, and production support. The role partners directly with enterprise customers, executive sponsors, and cross-functional teams to solve ambiguous problems, demonstrate prototypes, and standardize successful solutions. Regular domestic and occasional international travel is required.
Top Skills:
Api GatewaysAWSCi/CdDockerDuckdbFastmcpJavaKafkaKubernetesLangchainLanggraphMcpMtlsPgvectorPostgresPrivatelinkPythonRagTypescriptVpc Peering
Information Technology • Marketing Tech • Software • Design
Develop and deploy production machine learning systems for fall detection and health risk scoring. Build classical ML models, engineer features from radar and point-cloud data, contribute to pose estimation, and create Databricks training and inference pipelines. Own model evaluation and false-alarm reduction for a safety-critical system, monitor deployed models, and mentor team members while working in a fast-moving environment.
Top Skills:
Computer VisionDatabricksMlopsOnnxPoint-Cloud DataPythonRadar Signal ProcessingTime-Series ModelingXgboost
What you need to know about the Delhi Tech Scene
Delhi, India's capital city, is a place where tradition and progress co-exist. While Old Delhi is known for its rich history and bustling markets, New Delhi is defined by its modern architecture. It's clear the region places a strong emphasis on preserving its cultural heritage while embracing technological advancements, particularly in artificial intelligence, which plays a central role in shaping the city's tech landscape, fueled by investments in research and development.
