--- title: "Adjay Sagar S - Senior Data Scientist in Gurgaon | Tal" description: "Meet Adjay Sagar S, Senior Data Scientist @ Shiprocket | Transformers, BERT based in Gurgaon. AI stack includes antigravity, claude code, and claude." image: "https://tal.club/meet/adjay-sagar-s/share-image" --- # Adjay Sagar S on Tal ## Headline Senior Data Scientist @ Shiprocket \| Transformers, BERT ## About With over two years at Shiprocket, I lead machine learning and data science initiatives, focusing on voice AI and logistics intelligence\. Leveraging my expertise in Transformers, BERT, and Triton, I have built production\-grade TTS/STT systems, optimized inference speeds, and implemented scalable FastAPI services for high\-volume requests\. My work extends to advanced voice cloning and the deployment of robust systems on Triton Inference Server, utilizing PostgreSQL for efficient audit logs\. In logistics machine learning, I have contributed to enhancing delivery success rates through RTO prediction models and developed innovative address geocoding and intelligence pipelines for large\-scale logistics operations\. I am committed to advancing data\-driven solutions by combining cutting\-edge technologies with a collaborative approach to solving complex challenges in the AI and logistics domains\. ## Location Gurgaon ## AI stack antigravity claude code claude codex cursor ## Work experience Senior Data Scientist at Shiprocket, Jul 2024 \- Present · 2 yrs 2 mos, Gurugram, Haryana, India, Leading ML/DS development initiatives across voice AI and logistics intelligence: Voice AI Systems: \- Built production TTS/STT system from scratch \- Optimized inference to RTF < 0\.04 \(100s audio in 4s\) \- Designed FastAPI service handling 1000\+ req/min with caching \- Implemented voice cloning with 3\-step audio processing pipeline \- Deployed on Triton Inference Server with PostgreSQL for audit logs Logistics ML: \- Developed RTO prediction models improving delivery success rates \- Built address geocoding system for Delhi \(1\.5M\+ addresses indexed\) \- Created address intelligence pipeline \(standardization, validation\) Technical Stack: PyTorch, Triton, FastAPI, PostgreSQL, Redis, Docker, Ngrok, S3 Co\-Founder at CatchClues, Dec 2023 \- Jun 2024 · 7 mos, Gurugram, Haryana, India, Built a chat with your data application with RAG techniques and fine\-tuned versions of popular LLMs Built an Automated call recording audit system that transcribes and extracts key information from voice calls \- to improve customer satisfaction and identify potential abuses that might otherwise go unnoticed in manual audits Built a Vector\-based, low latency search system \(trained from scratch\) \- to smartly rank the most relevant products for search queries on e\-commerce platforms Built the ETL pipelines from scratch and also built a demand forecasting model using Gradient Boosting methods for an e\-commerce client Lead Analyst at Veera, Aug 2023 \- Dec 2023 · 5 mos, Gurugram, Haryana, India, Planned the events to be captured and designed the event data pipeline to track customer behaviour across the Veera Browser Android application Created the ETL pipelines and transformed the raw data for consumption Built Key KPI dashboards that helped product teams understand the highest retention levers and gaps in the Android app Senior Data Analyst at Blinkit, Jun 2022 \- Aug 2023 · 1 yr 3 mos, Gurugram, Haryana, India, Designed experiments for Focused targeting for fruits and vegetables campaign \- leading to better retentions using Simple Market Basket Analysis which led to increase in customer penetration \(from 33% to 38%\)\. Built a customer\-level new item propensity model basis Catboost model, which was used to predict if a customer would like a product even before they’ve seen it \- this helped us to optimise inventory across dark stores, also built ETL pipelines to generate the features needed for the machine learning model Identified POI \(Apartments\) wise purchase patterns of different categories and assisted the category managers in the planning of BTL campaigns at apartments and societies to increase brand awareness and run discounting Clustered search keywords using vector embeddings \(Sentence Transformers\) and then utilised it to identify bumps in demand for specific SKUs that helped in customising product catalogues for festivals \- overall Blinkit was able to achieve the highest ever sales during festivals with this approach Created an ops visibility dashboard from scratch that helped in optimising the supply chain operations at the inbound stage of Blinkit Warehouses \- This helped in better scheduling and quicker inwarding of products from vendors\. Operations Generalist at apna, Jan 2022 \- Jun 2022 · 6 mos, Chennai, Tamil Nadu, India, Built the ETL pipeline to track chat support metrics, earlier a manual process where KPIs were visible at the current\+1 date Extensively worked on improving chat user experience thru workflow design \+ aided it with automation \- DialogFlow\. Overall, this improved CSAT from 50% to 70% and increased the customer response Rate from 1\.5% to \~ 15% for CSAT surveys Built an SQL tool \- a heuristic\-based Employer Complaints fulfilment tool for RCAing issues in job postings\. That was later used by Sales \+ Ops \+ Misc\. Internal stakeholders Associate Product Manager at apna, Jan 2022 \- Jun 2022 · 6 mos, Developed an ETL pipeline to track chat support metrics, replacing a manual process that only surfaced KPIs at current\+1 date\. Redesigned chat user experience through workflow automation \(DialogFlow\), improving CSAT from 50% to 70% and raising survey response rate from 1\.5% to \~15%\. Created a heuristic\-based SQL tool for Employer Complaints RCA on job postings, adopted by Sales, Ops, and other internal stakeholders\. ## Education Bachelor's Degree, Chemical Engineering, Distinction Division, Birla Institute of Technology and Science, Pilani High School, Khaitan Public School [View the original profile](https://tal.club/meet/adjay-sagar-s) ```json { "@context": "https://schema.org", "@graph": [ { "@type": "Person", "@id": "https://tal.club/meet/adjay-sagar-s#person", "name": "Adjay Sagar S", "url": "https://tal.club/meet/adjay-sagar-s", "image": "https://tal.club/meet/adjay-sagar-s/profile-photo.jpg", "description": "Meet Adjay Sagar S, Senior Data Scientist @ Shiprocket | Transformers, BERT based in Gurgaon. AI stack includes antigravity, claude code, and claude.", "jobTitle": "Senior Data Scientist", "worksFor": { "@type": "Organization", "name": "Shiprocket" }, "hasOccupation": { "@type": "Occupation", "name": "Senior Data Scientist" }, "homeLocation": { "@type": "Place", "name": "Gurgaon" }, "alumniOf": [ { "@type": "EducationalOrganization", "name": "Birla Institute of Technology and Science, Pilani" }, { "@type": "EducationalOrganization", "name": "Khaitan Public School" } ], "knowsAbout": [ "antigravity", "claude code", "claude", "codex", "cursor" ], "mainEntityOfPage": { "@id": "https://tal.club/meet/adjay-sagar-s#webpage" } }, { "@type": "ProfilePage", "@id": "https://tal.club/meet/adjay-sagar-s#webpage", "url": "https://tal.club/meet/adjay-sagar-s", "name": "Adjay Sagar S on Tal", "description": "Meet Adjay Sagar S, Senior Data Scientist @ Shiprocket | Transformers, BERT based in Gurgaon. AI stack includes antigravity, claude code, and claude.", "dateModified": "2026-08-05T12:55:10.893Z", "isPartOf": { "@id": "https://tal.club/#website" }, "mainEntity": { "@id": "https://tal.club/meet/adjay-sagar-s#person" }, "primaryImageOfPage": { "@type": "ImageObject", "url": "https://tal.club/meet/adjay-sagar-s/profile-photo.jpg", "contentUrl": "https://tal.club/meet/adjay-sagar-s/profile-photo.jpg", "caption": "Adjay Sagar S profile photo on Tal" }, "inLanguage": "en-IN" } ] } ```