HemansAI

Data & AI Consulting, Development, AI Leadership Training.

Founder

85% of AI products and projects fail.

The two biggest challenges businesses face with AI:

1. Estimating and demonstrating AI value [VPs, CEOs, product managers, and decision-makers]
2. Lack of talent and skills [AI engineers and data professionals]

Most business leaders fall into the first category.

Even if you’re skilled at your job, your team might lack confidence that the AI solution will deliver the desired outcomes.

It’s frustrating, right? You see others succeeding with AI, but your team your team is stuck planning, fixing, or juggling disconnected tools.

AI is supposed to bring automation, sharper insights, and smarter decisions.

But when you try to implement it, you encounter chaos:

- Lack of clarity on ROI
- Endless troubleshooting instead of progress estimation
- Too many disconnected systems to demonstrate AI value

It’s easy to wonder if AI will ever deliver what it promises for your business and career.

Now, imagine a future where AI doesn’t just work—it works seamlessly in your organization.

Your data isn’t just stored—it powers real-time insights, precise forecasts, and simplified processes.

Picture an AI strategy that works without constant troubleshooting—a setup that scales and brings results.

I’m an AI consultant, here’s how I can help you:

- the AI leadership coaching

It’s about creating an AI path that’s right for your business or your career with real results, not just ideas.

Let’s connect. we’ll talk about AI strategies, products, and career opportunities.
connect@himanshuramchandani.co

Himanshu Ramchandani

Happiest Minds Technologies

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: 𝐆𝐞𝐧𝐀𝐈 𝐂𝐨-𝐏𝐢𝐥𝐨𝐭 𝐟𝐨𝐫 𝐀𝐮𝐝𝐢𝐭 𝐃𝐨𝐦𝐚𝐢𝐧 𝐐&𝐀'𝐬
𝐑𝐨𝐥𝐞: 𝐆𝐞𝐧𝐀𝐈 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐞𝐫
As a GenAI Developer, I worked on developing a GenAI Co-Pilot to assist users in the audit process. The primary objective was to provide quick and accurate answers to user queries, reducing manual searching time.

𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐢𝐞𝐬 𝐔𝐬𝐞𝐝
- Python
- Azure Open-AI GPT-4 LLM
- LlamaIndex as orchestration framework
- RAG technique for retrieving, augmenting, and generating responses
- Prompt techniques for structured answers and reducing hallucinations
- Streamlit: For interactive User Interface 

𝐈𝐦𝐩𝐚𝐜𝐭
The GenAI Co-Pilot helped users answer their questions, clear doubts, and reduced valuable time wasted in searching for correct answers. We achieved an accuracy of 86%, improved from 75% through optimization and unique implementation techniques.

Senior Data Scientist

Mindshare India

𝐏𝐫𝐨𝐣𝐞𝐜𝐭 – 𝐌𝐚𝐫𝐤𝐞𝐭 𝐌𝐢𝐱 𝐌𝐨𝐝𝐞𝐥𝐥𝐢𝐧𝐠- 𝐌𝐌𝐌			       
• 𝐎𝐛𝐣𝐞𝐜𝐭𝐢𝐯𝐞: Build a MMM model on the advertising spends data using regression for optimizing spends on media 
channels and analyze the impact of macroeconomic and other factors on the sales.
• Performed the data acquisition from GCP and client database, data preprocessing, data validation and 
analysis.
• Applied geometric adstock transformation on the media variables to introduce the advertising properties 
(e.g., decay, lag-lead) into the model. Used Facebook Prophet library to extract monthly 
trends.
• Built nested regression model to encounter issue of collinearity. Predicted store sales with 0.93 R2 score.
• Performed decomposition of the variables to understand the incremental effects of each media spends and 
external variables on sales, visualized using waterfall chart.
• Automated adstock hyperparameter tuning and deployed MMM nested model on local machine using 
streamlit.

Data Scientist

Enrichepro Software Pvt. Ltd. 

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: 𝐉𝐨𝐛 𝐭𝐢𝐭𝐥𝐞 𝐩𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐨𝐧 (𝐍𝐋𝐏 𝐂𝐥𝐚𝐬𝐬𝐢𝐟𝐢𝐜𝐚𝐭𝐢𝐨𝐧 𝐩𝐫𝐨𝐛𝐥𝐞𝐦)
• Given the features like Job description, domain info, job level, country, and other relevant features, performed NLP and feature engineering to use the text data information in predicting job title.
• Build and compared Logistic regression and tree-based ensemble models, selected XGBoost model which gave strong probability values for class labels. 
• Built end-to-end ML pipeline using MLflow Recipe, for model tracking and registering the validated model.

 𝐏𝐫𝐨𝐣𝐞𝐜𝐭 – 𝗖𝘂𝘀𝘁𝗼𝗺𝗲𝗿 𝘀𝗲𝗴𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻 𝘂𝘀𝗶𝗻𝗴 𝗰𝗹𝘂𝘀𝘁𝗲𝗿𝗶𝗻𝗴 𝗯𝘆 𝗶𝗻𝗰𝗼𝗿𝗽𝗼𝗿𝗮𝘁𝗶𝗻𝗴 𝗖𝗟𝗧𝗩 𝗮𝗻𝗱 𝗥𝗙𝗠 𝗮𝗻𝗮𝗹𝘆𝘀𝗶𝘀.
• 𝐎𝐛𝐣𝐞𝐜𝐭𝐢𝐯𝐞: Categorize the customers into High, Medium and Low value to drive the marketing efforts to maximize retention rate and acquire new customers.
• Performed data preprocessing and feature engineering to extract features like recency, frequency, monetary value (RFM), average order value, CLTV score, etc.
• Applied K-means clustering algorithm on the data with all the necessary features to segment the customers.
• Delivered results using dashboard and performed monitoring and retraining of the model.

ICICI Lombard

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: 𝐅𝐫𝐚𝐮𝐝 𝐜𝐥𝐚𝐢𝐦𝐬 𝐩𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐨𝐧 𝐚𝐧𝐝 𝐢𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐢𝐨𝐧 𝐨𝐟 𝐭𝐡𝐞 𝐟𝐥𝐚𝐠𝐠𝐢𝐧𝐠 𝐬𝐲𝐬𝐭𝐞𝐦 𝐢𝐧 𝐭𝐡𝐞 𝐬𝐨𝐟𝐭𝐰𝐚𝐫𝐞.
•	𝐎𝐛𝐣𝐞𝐜𝐭𝐢𝐯𝐞: Classification of the fraudulent claims and flagging them in the initial stages of claim process.
•	Developed understanding of business problem and contributed in designing project stages.
• Performed Data acquisition, preprocessing, EDA, and feature engineering to derive new features from 
existing. Used transfer learning using BERT to extract word vectors for text data i.e., cause of accident.
• Extracted the data using SQL and preprocessed the big data using parallel computing systems like Dask.
• Trained Logistic regression as base model and finalized Random Forest after comparing other models.
• Generalized the model by testing it against the test queries to make the model production ready.
• Contributed in end-to-end deployment in AWS Sagemaker, created endpoint and carried out API testing in Postman.

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: 𝐒𝐞𝐠𝐦𝐞𝐧𝐭𝐚𝐭𝐢𝐨𝐧 𝐨𝐟 𝐥𝐨𝐰, 𝐦𝐞𝐝𝐢𝐮𝐦 𝐚𝐧𝐝 𝐡𝐢𝐠𝐡 𝐥𝐢𝐚𝐛𝐢𝐥𝐢𝐭𝐲 𝐜𝐥𝐚𝐢𝐦𝐬 𝐮𝐬𝐢𝐧𝐠 𝐜𝐥𝐮𝐬𝐭𝐞𝐫𝐢𝐧𝐠.
•	𝐎𝐛𝐣𝐞𝐜𝐭𝐢𝐯𝐞: Segmenting the claims in low, medium and high liability claims to maintain equal workloads amongst surveyors as well as assigning high liability claims to external surveyors.
•	Developed understanding of business problem and performed data acquisition, data preprocessing and EDA.
• Used domain knowledge and feature engineering techniques to derive the new features from existing.
• Used K-means clustering to cluster the claims in low, medium and high liability claims.
• Developed a rule-based system with the help of IT team to maintain the workload within surveyors.

Data Science Analyst (CSM)

HDFC ERGO General Insurance

𝐏𝐫𝐨𝐣𝐞𝐜𝐭: 𝐅𝐫𝐚𝐮𝐝 𝐢𝐧𝐬𝐮𝐫𝐚𝐧𝐜𝐞 𝐜𝐥𝐚𝐢𝐦 𝐩𝐫𝐞𝐝𝐢𝐜𝐭𝐢𝐨𝐧 𝐚𝐧𝐝 𝐚𝐥𝐞𝐫𝐭𝐢𝐧𝐠.
• 𝐎𝐛𝐣𝐞𝐜𝐭𝐢𝐯𝐞: Classification of the fraudulent claims and integrating it with app that will send notification
alert.
• Performed data extraction from company’s database using MySql, stored and maintained updated data.
• Performed preprocessing like, removing duplicate and missing value rows etc., using Python and Excel.
• Completed basic EDA using Python and Tableau and presented the insights of the data before team.
• Performed other supporting tasks to build and productionize the model.

Claims Analyst (Executive)

-𝐸𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒𝑑 𝐷𝑎𝑡𝑎 𝑆𝑐𝑖𝑒𝑛𝑡𝑖𝑠𝑡 𝑤𝑖𝑡ℎ 𝑏𝑢𝑠𝑖𝑛𝑒𝑠𝑠 𝑎𝑐𝑢𝑚𝑒𝑛. 𝐻𝑒𝑙𝑝𝑖𝑛𝑔 𝑏𝑢𝑠𝑖𝑛𝑒𝑠𝑠𝑒𝑠 𝑡ℎ𝑟𝑖𝑣𝑒 𝑡ℎ𝑟𝑜𝑢𝑔ℎ 𝐷𝑎𝑡𝑎 𝑆𝑐𝑖𝑒𝑛𝑐𝑒 𝑤𝑖𝑡ℎ 𝑝𝑟𝑜𝑣𝑒𝑛 𝑡𝑟𝑎𝑐𝑘 𝑟𝑒𝑐𝑜𝑟𝑑 𝑜𝑓 𝑖𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑛𝑔 𝑎𝑛𝑑 𝑎𝑝𝑝𝑙𝑦𝑖𝑛𝑔 𝐷𝑎𝑡𝑎 𝑆𝑐𝑖𝑒𝑛𝑐𝑒 𝑡𝑒𝑐ℎ𝑛𝑖𝑞𝑢𝑒𝑠 𝑡𝑜 𝐵𝑢𝑠𝑖𝑛𝑒𝑠𝑠 𝑝𝑟𝑜𝑏𝑙𝑒𝑚𝑠.
-𝐸𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒 𝑖𝑛 𝑑𝑒𝑣𝑒𝑙𝑜𝑝𝑚𝑒𝑛𝑡 𝑎𝑛𝑑 𝑑𝑒𝑝𝑙𝑜𝑦𝑚𝑒𝑛𝑡 𝑜𝑓 𝑀𝑎𝑐ℎ𝑖𝑛𝑒 𝐿𝑒𝑎𝑟𝑛𝑖𝑛𝑔, 𝐷𝑒𝑒𝑝 𝐿𝑒𝑎𝑟𝑛𝑖𝑛𝑔 𝑎𝑛𝑑 𝑀𝑎𝑟𝑘𝑒𝑡 𝑀𝑖𝑥 𝑀𝑜𝑑𝑒𝑙𝑠 𝑢𝑠𝑖𝑛𝑔 𝑃𝑦𝑡ℎ𝑜𝑛 𝑜𝑛 𝑙𝑜𝑐𝑎𝑙 𝑚𝑎𝑐ℎ𝑖𝑛𝑒 𝑎𝑛𝑑 𝑐𝑙𝑜𝑢𝑑 𝑖𝑛𝑠𝑡𝑎𝑛𝑐𝑒𝑠.
-𝐵𝑒𝑙𝑖𝑒𝑣𝑒 𝑖𝑛 𝑡ℎ𝑒 𝑏𝑒𝑎𝑢𝑡𝑦 𝑜𝑓 𝑝𝑟𝑒𝑐𝑖𝑠𝑖𝑜𝑛, 𝑡ℎ𝑒 𝑝𝑜𝑤𝑒𝑟 𝑜𝑓 ℎ𝑎𝑟𝑑 𝑤𝑜𝑟𝑘, 𝑎𝑛𝑑 𝑡ℎ𝑒 𝑛𝑒𝑐𝑒𝑠𝑠𝑖𝑡𝑦 𝑜𝑓 𝑐𝑜𝑚𝑝𝑙𝑒𝑡𝑖𝑜𝑛.
-𝑆𝑝𝑒𝑐𝑖𝑎𝑙𝑡𝑖𝑒𝑠: 𝐼𝑑𝑒𝑎𝑡𝑖𝑜𝑛, 𝑒𝑥𝑒𝑐𝑢𝑡𝑖𝑜𝑛 𝑠𝑡𝑟𝑎𝑡𝑒𝑔𝑦, 𝑎𝑛𝑑 𝑑𝑟𝑖𝑣𝑖𝑛𝑔 𝑎 𝑝𝑟𝑜𝑑𝑢𝑐𝑡 𝑓𝑟𝑜𝑚 𝑠𝑡𝑎𝑟𝑡-𝑡𝑜-𝑒𝑛𝑑.
-𝑅𝑒𝑐𝑒𝑖𝑣𝑒𝑑 𝐴𝑤𝑎𝑟𝑑: 𝐺𝑟𝑜𝑢𝑝𝑀 𝐶ℎ𝑜𝑟𝑒𝑜𝑠'22 𝑎𝑤𝑎𝑟𝑑𝑠- 𝑁𝑒𝑤𝑏𝑖𝑒 𝐼𝑛 𝑇ℎ𝑒 𝑆𝑝𝑜𝑡𝑙𝑖𝑔ℎ𝑡.

taplio

The Battle

Himanshu Ramchandani

Akshay Bhasme

Followers

Posts

Engagements

Likes

Ready to grow your LinkedIn presence?