Sanketh Chebbi

Irrationally obsessed. Incurably curious.
Quantitative Orchestrator. Data Engineering. Applied AI.

Bangalore, KA, India

sanketh@duck.com • sanketh.in • GitHub • LinkedIn • X

Executive Summary

I'm a quantitative developer at True Beacon focused on maximizing shareholder value.
Operating at the intersection of data engineering and quantitative research, building battle-tested APIs and DAG-driven pipelines for large-scale market data. Designing and maintaining high-performance equity strategies with production-grade rigor. Tags: data, applied AI, buy-side-tech, financial-infrastructure.

Education

PES University Nov 2020 - Dec 2024
Bachelor's Degree in Mechanical Engineering

Work Experience

True Beacon Investment Advisors LLP May 2024 - Present
Quantitative Developer On-site
  • Built and productionized an in-house Order Management System supporting daily trading across portfolios managing over ₹500 Cr in AUM.
  • Designed and productionized the firm's flagship equity strategy with database-first research pipeline.
  • Architected end-to-end market data pipelines ingesting data from Bloomberg, ACE Equity into PostgreSQL, SQLite, and DuckDB.
  • Orchestrated all workflows using Prefect, making pipelines fault-tolerant, observable, and restart-safe.
  • Built a private, on-prem Generative AI system for fundamental research with retrieval-augmented analysis.
ISRO - ISTRAC Jan 2024 - May 2024
Scientist Engineer Intern On-site
  • Automated design of Back-up structure for Reflector Parabolic Antenna using Galapagos and Grasshopper.
  • Optimized for efficiency using Genetic Algorithm for less deflection and more stiffness.
  • Developed truss analysis software in Python using Stiffness Matrix Method.
  • Conducted EDA on departmental and project budgets for better resource management.
Mainstreet Marketplace Apr 2023 - Jun 2023
CRM Tech Lead Remote
  • Built a robust CRM platform for the team
  • Created a data site using React and Bootstrap with Shopify Storefront APIs
  • Unified platform for updating delivery and procurement status of products
Groww Apr 2022 - Sept 2022
SDE Intern On-site
  • Worked on iOS app using React Native
  • Implemented Dynamic Font Scaling in React Native
  • Migrated codebase from Redux to Redux Toolkit
  • Learned about React Native internals, state management, and JS code obfuscation

Projects

Agentic Equity Researcher : Multi-agent AI system for automated, auditable equity research workflows.
  • Designed and built an agentic orchestration framework to automate end-to-end equity research, spanning data ingestion, fundamental analysis, risk assessment, and report generation based on pre-set rules.
  • Advanced Citations: LLM-generated responses are enriched with precise, verifiable citations including source document names, page numbers, contextual text highlights, and extracted images. For RAG-based answers, users can interactively browse the cited documents directly within the response view and download PDFs with highlighted evidence.
  • Reduced analyst research turnaround time by ~60-70% for coverage updates while improving consistency and reproducibility of analysis across runs
  • System processed 10K+ documents across multiple equities with sub-second retrieval latency and controlled LLM inference costs via caching and model routing
UPIDigest : Unified expense tracking and AI-powered daily financial summaries.
  • Couldn't find a single piece of software to analyse gpay statements so I built one for myself (Goa Trip)
  • Built a web app to ingest, reconcile, and classify UPI and bank statements into a single expense view
  • Automatically matches transactions across sources and segregates categories and spending patterns
  • Generates a daily natural-language expense digest using an on-device LLM (using ollama with Mistral 7B)
Dataspear : Convert any SQL query into a fully parameterized, production-ready REST API.
  • Automatically exposes arbitrary SQL queries as REST endpoints with schema-aware parameterization
  • Built-in RBAC, request validation, and query safety controls to prevent misuse
  • Designed for internal analytics teams to ship data APIs without writing backend code
Self-Healing Data Pipeline Orchestrator : Config-driven ETL orchestration with AI-assisted recovery.
  • Built a custom workflow orchestrator for time-series ETL pipelines using declarative configuration
  • Implements automatic retries, checkpointing, and failure recovery without manual intervention
  • Uses AI-driven log analysis and anomaly detection to classify failures and suggest corrective actions
Banking Event Normalization Pipeline : Canonical transaction modeling across heterogeneous systems.
  • Designed and built a data pipeline to ingest raw transaction events from multiple banking sources
  • Normalized inconsistent schemas into a single canonical event model for downstream analytics and AI
  • Handled late-arriving events, schema evolution, and versioned transformations
PII-Aware Feature Store : Governance-first feature management for regulated AI systems.
  • Designed a feature store with explicit PII classification at the feature level
  • Enforced access controls and masking rules based on consumer identity and use case
  • Implemented detailed audit logging to support compliance and model governance

What do I bring to the table? I bring the table.

Things I believe in