AI Product Engineer
- Compensation
- ₹1.5M–₹2.5M / yr
- Location
- Hyderabad
- Work setup
- Hybrid
- Seniority
- Not specified
- Employment
- Full time
- Experience
- 5+ years
- Category
- Ai engineering
- Posted
Skills
Must have
- 3+ years building and shipping software, with at least 1–2 years building AI-powered applications in production (not just prototypes or coursework)
- Strong Python and modern full-stack skills (APIs, databases, front-end fundamentals); TypeScript is a plus
- Hands-on experience with applied AI: LLM APIs, agent frameworks, retrieval and structured-output patterns, and ML model integration — and the judgment to know when a simpler approach beats an AI one
- Evidence of speed: shipped products, side projects, or open-source work you can walk through — we hire on what you have built, not what you claim
- Data fluency: you can work with messy enterprise data, design pipelines, and debug why a model or integration is wrong
- Product sense: you ask what problem we are solving and who pays for it before you ask which model to use
- Comfort with ambiguity, enterprise clients, and production ownership — including fixing what you shipped
Nice to have
- Experience with forecasting or optimization (time series, operations research)
- AI governance, compliance, or audit-trail experience (NIST AI RMF, SOC 2, regulated industries)
- Cloud deployment (Azure/AWS/GCP), containers, CI/CD, and observability
- Prior work in a startup, GCC, or client-delivery environment
Full description
AI Product Engineer ACI InfotechLocation: Hyderabad / hybrid
Employment type: Full-time
Reports to: [Head of AI Products / Arq AI product leadership]
About ACI and Arq AI
ACI Infotech (founded 2006) is an enterprise technology firm serving Fortune 500 clients in data modernization, AI/ML implementation, cloud transformation, and ERP. Arq AI is ACI's applied-AI products platform a sister concern of ACI Infotech building vertical AI products such as ArqForecast (labor and demand forecasting), ArqGovernance (AI governance and compliance), and ArqSupport. Our positioning is applied AI and ML that ships to production, not demos.
The role
We are hiring an AI Product Engineer a builder who can take an AI application from idea to production at a rapid pace. You will own products end to end: understand the business problem, prototype fast, harden what works, and ship it to real users at enterprise clients. This is not a research role and not a ticket-taking role. If you need a perfect spec before you start, this is not the job.
What you will build
- Production AI applications: forecasting tools, AI assistants, workflow automation, and agent-based features for enterprise clients
- Prototypes in days that become production systems you decide what survives contact with real data and real users
- Full-stack delivery: APIs, data pipelines, front ends, and integrations with client systems
- Evaluation and guardrails: accuracy checks, failure modes, audit trails, and responsible-AI controls (ArqGovernance patterns) built in from the start
- Reusable product patterns: every deployment should make the next one faster — configuration over custom code
Must-haves
- 3+ years building and shipping software, with at least 1–2 years building AI-powered applications in production (not just prototypes or coursework)
- Strong Python and modern full-stack skills (APIs, databases, front-end fundamentals); TypeScript is a plus
- Hands-on experience with applied AI: LLM APIs, agent frameworks, retrieval and structured-output patterns, and ML model integration — and the judgment to know when a simpler approach beats an AI one
- Evidence of speed: shipped products, side projects, or open-source work you can walk through — we hire on what you have built, not what you claim
- Data fluency: you can work with messy enterprise data, design pipelines, and debug why a model or integration is wrong
- Product sense: you ask what problem we are solving and who pays for it before you ask which model to use
- Comfort with ambiguity, enterprise clients, and production ownership — including fixing what you shipped