Business Analyst - AI Platform

CLOUDLEX |

Location: Pune, India | Employment: Full time | Work mode: Work from office | Experience: 3 to 5 years

About CloudLex

CloudLex is a leading cloud based legal technology platform for litigation firms in the United States. Headquartered in New York, with engineering, product, and marketing in Pune, India. Reaching 300 professionals globally. Backed by prestigious venture capital and growth private equity, including Aligned Capital, Rosepark Advisors, and The Riverside Company.

CloudLex is a connected ecosystem: the platform that United States litigation firms run their entire practice on, an AI layer built into it, and managed services delivered inside the case file.

The Role: Business Analysis for the Future

AI now writes requirements documents, user stories, and specifications in minutes, and writes them well. That frees the Business Analyst to focus on the harder part: deciding what should be built, understanding the workflow deeply enough to know where it breaks, and judging whether what the platform produces is accurate enough to go in front of a lawyer. You will own a product area, decide what gets built in it, and define how the platform behaves.

Your Role

Use AI to produce the work

- Draft business requirements, functional specifications, user stories, and acceptance criteria using AI, then review and correct them before they go to engineering
- Build AI workflows that capture requirements and constraints from existing sources: support tickets, call recordings, user feedback, and current documentation
- Set up AI assisted research on competitors, industry trends, and workflow comparisons
- Maintain a prompt and template library the whole product team can reuse

Analyze the product with data
- Define what needs to be tracked in each workflow and work with engineering to instrument it
- Write SQL queries and prompts to pull usage data, and use AI to speed up the analysis




- Produce a monthly read on how your product area is actually being used
- Recommend what to build next, with the usage data behind the recommendation

Define how the platform behaves
- Write acceptance criteria for both standard features and AI driven features
- Build test sets that check whether AI output is accurate and complete
- Define confidence thresholds, what the system does when it is unsure, and where a human must approve before anything reaches a case file
- Work with engineering and QA through the sprint until the feature ships

What you will learn here
- Prompt engineering. How to get reliable, repeatable output from a model, and how to tell when the prompt is the problem rather than the model
- RAG, or Retrieval Augmented Generation. How the platform feeds a model the right case context, and what that means for the specifications you write
- AI agents. How multi-step AI workflows are built, where they fail, and where a human checkpoint belongs
- Data storytelling. Turning usage data into an argument that changes what the team builds
- Working Python. Enough to pull data, run your own analysis, and build small tools without waiting for engineering

What we need from you
- 3 to 5 years as a Business Analyst, Product Analyst, or Product Manager, preferably in B2B SaaS
- Daily hands on use of AI tools such as Claude, ChatGPT, or Gemini, with a clear view of where they fail
- SQL or equivalent. You should be able to pull data and reach a conclusion from it
- Solid written English. This role produces documents that American teams act on
- An example of a time you changed what the team decided to build, and were right

Please send one work sample with your application: a specification you wrote, an analysis that changed a decision, a test or evaluation framework, or something you built using AI tools. We read these, and they count for more than the resume.

Practical

- Pune, work from office
- Some overlap with United States hours is part of the role, and we would rather tell you now than after you join

JD

📌 Business Analyst (Pune)
🏢 CloudLex - The Legal Cloud
📍 Pune