Product Manager - AI Native

Clickpost; Bengaluru; On-site; INR 1,500,000 - 2,300,000; Full time

Job summary

ClickPost is seeking an AI-native Product Manager to own AI products and drive agent workflows. You will leverage AI coding tools to analyze data, build simulations, and prototype agent workflows, turning repetitive tasks into reusable tools.

In this role, you will own outcomes end-to-end, write clear PRDs, and collaborate across Engineering, Data Science, Customer Success, and Sales to deliver measurable improvements in delivery optimization and post-purchase intelligence.

Qualifications

• 3 to 5 years in product management, preferably in B2B SaaS.
• Proficient with SQL and Excel at scale.
• Experience using AI coding tools and building simulations.
• Able to read API docs and logs and discuss trade-offs with engineers.
• Proven track record shipping features with measurable impact.

Responsibilities

• Own AI products and build agent workflows.
• Use AI to analyze data, run simulations, and prototype sequences.
• Define PRDs and track metrics weekly.
• Collaborate with Engineering, Data Science, Customer Success, and Sales.
• Ship AI-powered features to real users and measure impact.

Skills

AI tooling

SQL

Python

Data analysis

Product management

Tools

Claude Code

Cursor

Job description

Department: Product

Experience: 3 to 5 YOE

About us :

Since 2017, ClickPost has been the post-purchase intelligence platform that turns logistics complexity into a competitive advantage for the world's leading retailers, including Walmart, Nykaa, Mars, Adidas, Bluestone, GIVA, Kapiva, and 450+ brands across India, Southeast Asia, the Middle East, and North America. Through a single API, our platform manages tracking, exceptions, and returns across 600+ pre-integrated carriers, with AI allocating every shipment to the optimal carrier and resolving delivery issues before orders fail.

And we're not standing still. We're actively building the agentic layer for post-purchase, from AI voice agents like Parth to a growing suite of agents that resolve delivery exceptions, confirm orders, and recover lost sales. This is real, in-production AI, live with customers today.

We process 50M+ shipments every month, helping brands cut failed deliveries and "where-is-my-order" queries while converting more returns into exchanges.

We're one of the rare companies operating at the intersection of cutting-edge AI, enormous scale, and a brand respected across the global SaaS and e-commerce world. Our core values of transparency, continuous learning, ownership, and speed, plus a team that includes several ex-startup founders, make this a place that's both fast-moving and rich with opportunities to learn. Our vision is to become the global standard for post-purchase intelligence, where logistics complexity never limits a brand's growth.

About the Role:

We help e-commerce brands ship millions of orders through many delivery partners. Behind every order are a few simple questions. Which delivery partners can reach this customer? Which one is the best choice for this order? When will it actually arrive? And what happens when something goes wrong?

You'll own the products that answer those questions, along with the AI agents we're building on top of them.

We're looking for an AI-native PM. You'll own AI products, and you'll also use AI for most of your own work. You'll use AI coding tools like Claude Code or Cursor to analyze data, build simulations, prototype agent workflows, investigate issues, and turn your repeated tasks into reusable tools.

Something that takes a traditional PM a week with an analyst and an engineer should take you an afternoon.

What you will do:

1.Delivery coverage

• How brands set up which delivery partners can serve which locations.
• Making setup faster for new brands and reducing configuration errors and support tickets.

Outcome : Brands go live faster, and fewer orders fail because no suitable delivery partner was available.

2. Delivery partner selection and recommendations

• The rules and rankings that decide which delivery partner handles each order.
• Balancing cost, speed, delivery success, returns and partner capacity for each brand.

Outcome : Lower cost per delivered order, higher delivery success, and brands that can see why each partner was chosen.

3. Simulation

• Running what-if simulations before anything changes in production, like "what happens to cost and delivery time if we move 15% of orders from one partner to another?"
• Turning the results into clear recommendations that brands can act on.

Outcome : Every major change is backed by a simulation before it goes live.

4. Delivery date promise

• How accurate the delivery date shown to shoppers is, how often we miss it, and how that affects the brand's sales and customer experience.

Outcome : Shoppers see delivery dates they can trust, and brands see fewer complaints about late orders.

5. AI agents

• Defining and shipping AI agents that recommend strategies, explain performance changes, and answer customer questions on their own.
• Owning the evaluation: when the agent is right, when it's wrong, and how we know which.

Outcome : Agents that resolve a measurable share of questions end to end, with answers people can trust and act on

6. Data analysis and customer issue resolution

• Digging into large datasets to answer questions like "why did this partner's share of orders jump last week?" or "why did delivery success drop in this region?"
• Owning customer escalations from the moment they're reported until they're closed, and turning the patterns you see into product fixes.

Outcome : Faster answers for customers, and fewer of the same issues coming back.

How you'll work:

• You build things yourself. Use AI coding tools to write SQL, Python scripts and simulations, and build working prototypes without waiting for engineering bandwidth.
• You automate what you repeat . If you investigate the same type of issue twice, the third time it should be a reusable script, tool or agent workflow.
• You turn your knowledge into something AI can use . Write the guides, knowledge bases and prompts that our AI agents and internal teams rely on.
• You test AI properly . Define evaluations, test against real cases, find the failure modes, and know when an agent is confidently wrong.
• You prototype before you write the spec. Show engineering a working prototype instead of a 10-page document whenever you can.
• You own outcomes end-to-end. Write clear PRDs, then stay involved through build, testing, rollout and adoption. Track your products' metrics every week and act when they drift.
• You work across teams. Work directly with Engineering, Data Science, Customer Success, Sales and customers.

What success looks like: First 90 Days

• First 30 days: You understand how our core products work end-to-end, you've investigated real customer cases, and you've resolved your first few escalations on your own.
• By 90 days: You own one or two product areas outright, you've shipped at least one meaningful improvement, and you've delivered at least one simulation-backed recommendation to a customer.
• By 6 months: Your products' metrics (delivery date accuracy, recommendation quality, escalation volume, agent accuracy) have measurably improved, and you've automated a meaningful share of your own recurring work.

What we're looking for:

Must have:

• 3 to 5 years in product management, preferably in B2B SaaS, logistics, e-commerce, fintech or marketplaces.
• You use AI daily, in depth . You regularly use AI coding tools and LLMs to build and analyse things, not only to write emails. You can show us what you've built.
• Strong with data. SQL and Excel at scale, plus enough Python, with AI assistance, to run your own analysis and simulations.
• Comfortable with technical detail. You can read API docs and logs, and discuss trade-offs with engineers.
• You understand how LLMs behave . You know about context, tool use, hallucinations, evaluations, and when to use an LLM versus fixed rules.
• A record of shipping. Concrete examples of features you took to production, with measurable impact.
• High ownership and urgency . You close loops, unblock other teams, elevate early with a proposed fix, and don't need follow-ups.
• Structured thinking . You can take a vague customer complaint and break it down into root causes.

Good to have:

• Experience in logistics, supply chain or e-commerce operations.
• Experience with recommendation, ranking, optimisation or forecasting products.
• You've built your own AI agents, tools or automations.
• You've shipped an LLM-powered feature to real users and handled what broke afterwards.
• A GitHub profile, a portfolio, or side projects that show you build things.

What You Can Expect:

• Ownership and autonomy: You'll own and drive problems end-to-end, with no micro-management.
• Culture: We value speed and expect you to move fast and stay updated with the latest in ML and AI. People grow extremely fast at ClickPost when they focus on impact and take initiative.
• Our team thrives on transparency.
• Compensation: Competitive salary plus performance bonus.