Pratiksha S. profile photo 1 on Tal

Pratiksha S.

Software Engineer @ Edelweiss | Backend & Data Platform | Java, Python, Kafka, Spring Boot | Ex-BNY, Morgan Stanley | GSOC’24 | LFX’24 | C4GT’24 | SIH’23 Finalist | VJTI CS’24

Mumbai

2+ years of work experience

Edelweiss Global Markets
Oct 20251 yrPresent

Mumbai Metropolitan Region

- Built Fintasker, an in-house orchestrator that replaced Airflow firm-wide, running 1,000+ production pipelines across a 20-hour trading window through a distributed, fault-tolerant architecture with independent task execution. - Own the market-data ingestion layer for 9 global exchanges (US, Korea, Taiwan, Japan) across cash, futures, and options, streaming via Kafka with ~98% availability during market hours. - Designed Cerberus, a data-quality framework spanning 10+ categories and 250+ rules, cutting downstream anomalies by 75%. - Launched Astra, a trader analytics platform (multi-source news, FII/DII flows, gainers/losers, corporate actions, IPOs, index heatmap), adopted by 200+ traders firm-wide. - Scaled Data Mart APIs using Spring Boot and distributed ClickHouse; built ETL pipelines onboarding LSEG and Databento historical data at terabyte scale.
Nov 20242 mosDec 2024

• Designed and developed the website, integrating backend systems and streamlining access to judicial information and services. • Redesigned the UI for enhanced accessibility, responsiveness, andmodern user experience.
Bank of New York
Jul 20241 y 3 moOct 2025

Pune District, Maharashtra, India

- Built the prompt-validation and evaluation service (Spring Boot) gating every LLM feature on Eliza, using ground-truth datasets and audit logging to cut manual validation time from 1 week to 1 day. - Built FastAPI-based LLM agents automating multi-source credit-risk document generation, reducing prep time from ~40 days to 3-5 days. - Shipped custom JupyterLab extensions for Eliza, BNY's internal AI platform, streamlining experimentation workflows.
The Linux Foundation
Jun 20246 mosNov 2024

Mumbai Metropolitan Region

- Contributed to improving Hyperledger Caliper documentation, focusing on clarity, accessibility, and consistency. - Migrated Caliper documentation to the standard Hyperledger documentation template, ensuring a unified experience across projects.
Google Summer of Code
May 20247 mosNov 2024

Mumbai, Maharashtra, India

- Developed reusable web components using LitElement aligned with GA4GH DRS API specifications. - Built a secure and scalable backend using DRS-Filer and MinIO for cloud-based data storage. - Upgraded DRS-Filer to the latest DRS API version, ensuring interoperability within the ELIXIR Cloud Component ecosystem.
GitHub
@psankhe28

Pratiksha’s AI Stack

antigravity
antigravity
claude code
claude code
claude
claude
codex
codex
cursor
cursor
github copilot
github copilot

Education

Veermata Jijabai Technological Institute (VJTI)

Bachelor of Technology, Computer Engineering

Veermata Jijabai Technological Institute (VJTI)

20213 yrs2024

Join Pratiksha and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Pratiksha S. on Tal

Headline

Software Engineer @ Edelweiss | Backend & Data Platform | Java, Python, Kafka, Spring Boot | Ex-BNY, Morgan Stanley | GSOC’24 | LFX’24 | C4GT’24 | SIH’23 Finalist | VJTI CS’24

About

Software engineer building high-throughput, fault-tolerant backend and data systems for real-time financial markets. At Edelweiss Global Markets, I built Fintasker, an in-house orchestrator that replaced Airflow team-wide, running 1,000+ production pipelines across a 20-hour trading window. I own market-data ingestion across 9 global exchanges (US, Korea, Taiwan, Japan) via Kafka, and designed Cerberus, a data-quality framework that cut downstream anomalies by 75%. Previously at Bank of New York's AI Hub, I built the prompt-validation and evaluation infrastructure gating every LLM feature on the bank's internal AI platform, and shipped FastAPI-based agents that automated credit-risk document generation, cutting prep time from ~40 days to 3-5. I work primarily in Java, Python, and Spring Boot, with hands-on experience across Kafka, ClickHouse, PostgreSQL, and distributed system design. Outside work, I contribute to open source — Google Summer of Code (ELIXIR Cloud & AAI), Linux Foundation, and Code for GovTech — and I'm a B.Tech graduate from VJTI (CGPA 9.22/10). Open to backend and data engineering roles where I can work on systems at scale.

Location

Mumbai

AI stack

antigravity

claude code

claude

codex

cursor

github copilot

Work experience

Assistant Manager — Software Development Engineer, Data Platform at Edelweiss Global Markets, Oct 2025 - Present · 1 yr, Mumbai Metropolitan Region, - Built Fintasker, an in-house orchestrator that replaced Airflow firm-wide, running 1,000+ production pipelines across a 20-hour trading window through a distributed, fault-tolerant architecture with independent task execution. - Own the market-data ingestion layer for 9 global exchanges (US, Korea, Taiwan, Japan) across cash, futures, and options, streaming via Kafka with ~98% availability during market hours. - Designed Cerberus, a data-quality framework spanning 10+ categories and 250+ rules, cutting downstream anomalies by 75%. - Launched Astra, a trader analytics platform (multi-source news, FII/DII flows, gainers/losers, corporate actions, IPOs, index heatmap), adopted by 200+ traders firm-wide. - Scaled Data Mart APIs using Spring Boot and distributed ClickHouse; built ETL pipelines onboarding LSEG and Databento historical data at terabyte scale.

Web Developer at Kerala OnCourts, Nov 2024 - Dec 2024 · 2 mos, • Designed and developed the website, integrating backend systems and streamlining access to judicial information and services. • Redesigned the UI for enhanced accessibility, responsiveness, andmodern user experience.

Software Engineer at Bank of New York, Jul 2024 - Oct 2025 · 1 y 3 mo, Pune District, Maharashtra, India, - Built the prompt-validation and evaluation service (Spring Boot) gating every LLM feature on Eliza, using ground-truth datasets and audit logging to cut manual validation time from 1 week to 1 day. - Built FastAPI-based LLM agents automating multi-source credit-risk document generation, reducing prep time from ~40 days to 3-5 days. - Shipped custom JupyterLab extensions for Eliza, BNY's internal AI platform, streamlining experimentation workflows.

Open Source Developer at The Linux Foundation, Jun 2024 - Nov 2024 · 6 mos, Mumbai Metropolitan Region, - Contributed to improving Hyperledger Caliper documentation, focusing on clarity, accessibility, and consistency. - Migrated Caliper documentation to the standard Hyperledger documentation template, ensuring a unified experience across projects.

Open Source Developer at Google Summer of Code, May 2024 - Nov 2024 · 7 mos, Mumbai, Maharashtra, India, - Developed reusable web components using LitElement aligned with GA4GH DRS API specifications. - Built a secure and scalable backend using DRS-Filer and MinIO for cloud-based data storage. - Upgraded DRS-Filer to the latest DRS API version, ensuring interoperability within the ELIXIR Cloud Component ecosystem.

Education

Bachelor of Technology, Computer Engineering, Veermata Jijabai Technological Institute (VJTI)

Public links

github: https://github.com/psankhe28