AI Architect

About the Company

At Horizon17 Technology and Sustainability Pvt Ltd, we harness the power of technology to accelerate global progress towards sustainability. Founded by a team of forward-thinking sustainability and technology experts, we are committed to advancing the Environmental, Social, and Governance (ESG) agenda by integrating cutting-edge innovations with purposeful action. Our aim is to infuse transformative technologies like Blockchain, Artificial Intelligence, and Advanced Data Analytics into the framework of the United Nations Sustainable Development Goals (SDGs).

Company URL: https://horizon17ww.com

LinkedIn: https://www.linkedin.com/company/horizon17-tspl

Job Title: AI Architect

Location: Gurgaon / Remote

About the Role

We're seeking a hands-on AI Architect to design and build an internal platform providing full visibility into AI usage across our organization — covering usage, cost, model routing, performance, and quality. This is a builder role with end-to-end ownership: you'll architect the system, select tools, code it from scratch, and ensure it's production-ready for our team to maintain and scale.

Project Context

Our AI usage spans multiple LLM providers, agent frameworks, and observability stacks, leading to surprise costs and fragmented insights. You'll deliver a unified platform that ingests events, attributes them to teams/use-cases, and provides dashboards, alerts, and agentic reporting for proactive management.

Key Responsibilities

Architecture & Design

• Create a comprehensive architecture document (data ingestion, storage, agents, dashboards, security).
• Define tech stack with trade-off analysis (build vs. buy, open-source vs. vendor).
• Set targets for cost, latency, and reliability.

Core Platform Build

• Ingest LLM events (tokens, latency, cost, model, user, team, use-case) from existing workloads.
• Build normalized data store and query layer for reporting/analysis.
• Develop dashboards for cost, usage, and quality by team/model/use-case.
• Add alerting for cost spikes, errors, and anomalies.

Agentic Layer

• Implement natural-language querying for AI usage/cost insights.
• Build scheduled agents for weekly/monthly summaries, anomaly reports, and recommendations.
• Add guardrails, evaluations, and human-in-the-loop for reliability.

Required Qualifications

• 6–8 years in software engineering; 5+ years building production AI/LLM systems.
• End-to-end design/shipping of non-trivial agentic systems.
• Production ops in AWS, Azure, or GCP (networking, storage, IAM, managed services).
• Expert Python (type hints, testing, async, packaging).
• Experience with Agent frameworks: LangGraph, LangChain, CrewAI, AutoGen, or equivalent.
• Experience with LLM observability: Langfuse, Helicone, OpenTelemetry for LLMs, etc.
• Data: PostgreSQL + analytical stores; event data modeling