Product Manager

Mstack AI; Bengaluru Urban; On-site; INR 1,200,000 - 2,000,000; Full time

Job summary

Mstack AI, Bengaluru-based tech-enabled chemicals company, seeks a product-focused professional to lead AI-driven automation and workflows. You will engage with sales, procurement, logistics and finance to define problems, prioritize solutions and ship AI-enabled features that reduce manual effort.

You will own PRD writing, wireframing in Figma, and coordinating with engineering through sprints to ensure successful releases. Strong AI tooling and data analysis skills are essential.

Qualifications

• 1-3 years of product management experience or PM internship leading to ownership of a product.
• Has shipped at least one product or feature end to end with measurable results.
• Has built something real with AI: agent, automation, internal tool or AI feature.
• Hands-on with modern AI tools and uses them daily.
• Familiar with LLM concepts: prompting, context, retrieval, tool calling, latency and cost.

Responsibilities

• Conduct user research with sales, procurement, logistics and finance users to derive problem statements.
• Size problems using data (time lost, cost, errors, revenue at risk) and decide what to solve first.
• Explore options and pick simple solutions that move metrics (software, AI agents, automation, or process change).
• Write crisp PRDs with goals, flows, edge cases, metrics and acceptance criteria for engineering.
• Create wireframes or AI-based prototypes in Figma and review usability.
• Collaborate with engineering through sprints and own release quality.
• Plan rollout, train users, publish release notes and track adoption.
• Own one or more product areas end to end and report metrics to leadership.
• Identify AI opportunities and design agentic workflows for email, voice and documents.
• Prototype fast with tools like Claude Code, Cursor, Lovable, n8n or Zapier.
• Draft PRDs covering prompts, context, evaluation sets, guardrails and human-in-the-loop.
• Set up evals/monitoring to know when an agent is right, wrong, or drifting.
• Share learnings to raise the team’s AI bar.

Skills

Product management

User research

PRD writing

AI literacy

Data analysis

Communication

Ownership

Education

Engineering degree

Tools

Claude Code

Cursor

Lovable

n8n

Zapier

Job description

Mstack is a tech-enabled specialty chemicals company that runs global sourcing, sales, logistics and finance on its own platforms. The Product team builds and runs SalesStack, ProcureStack and LogiStack, and integrates them with our ERP (Microsoft Dynamics Business Central).

We are now moving manual, high-volume work - supplier and customer emails, follow-up calls, document generation, invoice posting - to AI agents and automated workflows. This role sits at the centre of that shift.

What you\'ll do: core product management

• User research: spend time with sales, procurement, logistics and finance users; run interviews, shadow workflows and turn what you see into clear problem statements.
• Problem definition: size the problem with data (time lost, cost, error rates, revenue at risk) and decide what is worth solving now.
• Solutioning: explore options, weigh trade-offs, and pick the simplest solution that moves the metric - software, AI agent, automation or process change.
• PRDs: write crisp PRDs with goals, non-goals, user flows, edge cases, success metrics and acceptance criteria that engineering can build from.
• UI/UX: create wireframes and flows in Figma (or prototype directly with AI tools) and review designs for usability.
• Delivery: work with engineering through sprints, unblock decisions, groom the backlog and own release quality.
• GTM and adoption: plan rollout, train users, write release notes and track adoption until the feature is actually used.
• Ownership: own one or more product areas end to end, with metrics you report on to leadership.

What you\'ll do: AI, agents and automation

• Spot AI opportunities: find repetitive, rules-heavy or communication-heavy work across teams and judge where AI is the right tool and where it isn\'t.
• Design agentic workflows: define what an agent does, which tools and data it uses, where a human approves, and how it fails safely — for email, voice and document-based workflows.
• Prototype fast: build working prototypes and internal tools yourself using tools like Claude Code, Cursor, Lovable, n8n or Zapier before asking engineering to build at scale.
• Spec AI features well: write PRDs that cover prompts, context, evaluation sets, accuracy targets, guardrails, cost per run and human-in-the-loop design.
• Measure AI quality: set up evals and monitoring so we know when an agent is right, wrong or drifting.
• Raise the team\'s AI bar: share what works, help other teams use AI in their daily work, and keep up with new models and tools.

Must-haves

• 1-3 years of product management experience (PM, APM or a PM internship that led to owning a product).
• Has shipped at least one product or feature end to end, with measurable results you can explain.
• Has built something real with AI - an agent, an automation, an internal tool or an AI feature in production - and can walk through how it works.
• Hands-on with modern AI tools (LLM APIs or chat tools, AI coding tools, no-code automation platforms) and uses them in daily work.
• Working understanding of LLM concepts: prompting, context and retrieval, tool calling, hallucinations, evaluation, latency and cost.
• Comfortable with data: SQL basics, funnels, dashboards, and A/B or before-after analysis.
• Clear written communication - PRDs, updates and decisions that others can act on without a meeting.
• High ownership and comfort with ambiguity in a fast-moving team.

Nice-to-haves

• Engineering degree or a technical background; can read code and API docs.
• Experience with B2B, operations-heavy, supply chain, logistics or fintech products.
• Built voice agents, AI IVRs, email agents or document-extraction (OCR/LLM) workflows.
• Experience with ERP or CRM integrations (e.g., Business Central, Zoho, Salesforce).
• Familiarity with agent frameworks or orchestration tools (e.g., LangGraph, n8n, MCP, Claude Agent SDK).
• A public portfolio: side projects, GitHub, blog or demos of things you\'ve built.

What success looks like:

30 days

• Learned the business and our platforms; shadowed users in at least 3 teams; shipped one small improvement or automation.

60 days

• Owned a product area; written and shipped at least one PRD; prototyped one AI workflow and validated it with real users.

90 days

• Launched an AI agent or automation to production with a clear metric (hours saved, cost cut, cycle time or error rate) and a plan to scale it.