Sanyam Agrawal profile photo 1 on Tal

Sanyam Agrawal

LLM Backend Engineer | FastAPI | Async Systems | RAG, LangGraph, OpenAI | Scaling AI APIs to 10K+ req/day

Greater Bengaluru Area

4+ years of work experience

Infosys
Sep 20233 yrsPresent

• Built and scaled backend APIs handling 10K+ requests/day with FastAPI and async Python, achieving 99.2% uptime. • Reduced average API latency from 220ms → 140ms (36% improvement) by optimizing async I/O, batching, and connection pooling. • Improved P95 latency by 42% through parallel processing and non-blocking external API calls. • Designed structured logging and monitoring pipelines, reducing production debugging time by 40% and MTTR by 35%. • Implemented retry and circuit-breaker mechanisms, increasing request success rate from 91% → 98.5% under failure scenarios. • Reduced system error rates by 45% via improved validation, exception handling, and fallback logic. • Optimized database/API interactions, reducing redundant calls by 30% and improving throughput by 25%. • Led integration of external services/APIs, achieving 95%+ success rate across third-party dependencies. • Introduced caching strategies (Redis), reducing response time for repeated queries by 50%+ and lowering compute cost. • Wrote unit and integration tests using Pytest, improving code coverage to 85%+ and reducing regression issues by 30%.
Infosys
Sep 20232 yrs 8 mosApr 2026

Bengaluru

• Built and scaled backend APIs handling 10K+ requests/day, achieving 99.2% uptime • Reduced latency from 220ms → 140ms (36%) using async I/O and batching • Improved P95 latency by 42% via parallel processing • Increased success rate from 91% → 98.5% using retry + circuit breakers • Reduced error rates by 45% with better validation and fallback logic • Cut debugging time by 40% with structured logging • Improved throughput by 25% and reduced redundant calls by 30% • Implemented Redis caching → 50%+ faster responses • Achieved 85%+ test coverage using Pytest
Danske IT and Support Services India Pvt Ltd
Jul 20221 yr 2 mosAug 2023

Bengaluru

Worked on IMA (Information Management Architecture), an enterprise metadata platform covering lineage, technical, and business metadata across systems. • Built and maintained the platform, improving traceability across critical datasets • Designed data models and optimized SQL stored procedures, reducing query time by 25%+ • Developed backend services using Django, including APIs, data processing, and authentication flows • Integrated Azure Active Directory (LDAP) for secure, role-based access for 100+ users • Deployed and configured infrastructure using Cloud2 and Linux, including servers, endpoints, and firewall rules • Set up development environments and Git/GitHub workflows, improving collaboration and reducing onboarding time by 30–40% • Built frontend using Django templates and HTML, and managed SSL certification and deployment
Danske Bank
Jul 20221 yr 2 mosAug 2023

• Built and maintained IMA, an enterprise metadata platform covering lineage, technical, and business metadata across systems, improving traceability of critical datasets. • Designed data models and optimized SQL stored procedures, reducing query time by 25%+ and improving data retrieval efficiency. • Developed backend using Django, handling APIs, data processing, and authentication flows; integrated Azure AD (LDAP) for secure, role-based access for 100+ users. • Deployed and configured infrastructure using Cloud2 and Linux (servers, endpoints, firewall rules), enabling secure multi-environment access. • Set up development environments and Git/GitHub workflows, improving collaboration and reducing onboarding time by 30–40%. • Built frontend using Django templates/HTML, managed SSL certification and deployment, and drove adoption through documentation and onboarding of internal teams.
Danske IT and Support Services India Pvt Ltd
Jan 20227 mosJul 2022

Bengaluru, Karnataka, India

Sanyam’s AI Stack

langchain
langchain
langgraph
langgraph
langsmith
langsmith
notion
notion
claudecode
claudecode
githubcopilot
githubcopilot
ChatGPT
ChatGPT
Microsoft Copilot
Microsoft Copilot
Wispr Flow
Wispr Flow

Education

Vellore Institute of Technology

Bachelor of Technology, Computer Science and Engineering

Vellore Institute of Technology

20184 yrs2022
Delhi Public School - R. K. Puram

Delhi Public School - R. K. Puram

20152 yrs2017

Join Sanyam and 10k+ other professionals on

Tal Boss

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

Download Tal Boss

Sanyam Agrawal on Tal

Headline

LLM Backend Engineer | FastAPI | Async Systems | RAG, LangGraph, OpenAI | Scaling AI APIs to 10K+ req/day

About

Backend engineer focused on building production-grade LLM systems with strong emphasis on reliability, latency optimization, and cost efficiency. I specialize in designing and scaling AI-powered backend systems, including multi-model LLM integrations, agent workflows, and retrieval-based architectures. Interested in: LLM infrastructure, agent systems, and high-performance backend engineering.

Location

Greater Bengaluru Area

AI stack

langchain

langgraph

langsmith

notion

claudecode

githubcopilot

ChatGPT

Microsoft Copilot

Wispr Flow

Work experience

Software Engineer at Infosys, Sep 2023 - Present · 3 yrs, • Built and scaled backend APIs handling 10K+ requests/day with FastAPI and async Python, achieving 99.2% uptime. • Reduced average API latency from 220ms → 140ms (36% improvement) by optimizing async I/O, batching, and connection pooling. • Improved P95 latency by 42% through parallel processing and non-blocking external API calls. • Designed structured logging and monitoring pipelines, reducing production debugging time by 40% and MTTR by 35%. • Implemented retry and circuit-breaker mechanisms, increasing request success rate from 91% → 98.5% under failure scenarios. • Reduced system error rates by 45% via improved validation, exception handling, and fallback logic. • Optimized database/API interactions, reducing redundant calls by 30% and improving throughput by 25%. • Led integration of external services/APIs, achieving 95%+ success rate across third-party dependencies. • Introduced caching strategies (Redis), reducing response time for repeated queries by 50%+ and lowering compute cost. • Wrote unit and integration tests using Pytest, improving code coverage to 85%+ and reducing regression issues by 30%.

Technology Analyst at Infosys, Sep 2023 - Apr 2026 · 2 yrs 8 mos, Bengaluru, • Built and scaled backend APIs handling 10K+ requests/day, achieving 99.2% uptime • Reduced latency from 220ms → 140ms (36%) using async I/O and batching • Improved P95 latency by 42% via parallel processing • Increased success rate from 91% → 98.5% using retry + circuit breakers • Reduced error rates by 45% with better validation and fallback logic • Cut debugging time by 40% with structured logging • Improved throughput by 25% and reduced redundant calls by 30% • Implemented Redis caching → 50%+ faster responses • Achieved 85%+ test coverage using Pytest

Software Engineer at Danske IT and Support Services India Pvt Ltd, Jul 2022 - Aug 2023 · 1 yr 2 mos, Bengaluru, Worked on IMA (Information Management Architecture), an enterprise metadata platform covering lineage, technical, and business metadata across systems. • Built and maintained the platform, improving traceability across critical datasets • Designed data models and optimized SQL stored procedures, reducing query time by 25%+ • Developed backend services using Django, including APIs, data processing, and authentication flows • Integrated Azure Active Directory (LDAP) for secure, role-based access for 100+ users • Deployed and configured infrastructure using Cloud2 and Linux, including servers, endpoints, and firewall rules • Set up development environments and Git/GitHub workflows, improving collaboration and reducing onboarding time by 30–40% • Built frontend using Django templates and HTML, and managed SSL certification and deployment

Software Engineer at Danske Bank, Jul 2022 - Aug 2023 · 1 yr 2 mos, • Built and maintained IMA, an enterprise metadata platform covering lineage, technical, and business metadata across systems, improving traceability of critical datasets. • Designed data models and optimized SQL stored procedures, reducing query time by 25%+ and improving data retrieval efficiency. • Developed backend using Django, handling APIs, data processing, and authentication flows; integrated Azure AD (LDAP) for secure, role-based access for 100+ users. • Deployed and configured infrastructure using Cloud2 and Linux (servers, endpoints, firewall rules), enabling secure multi-environment access. • Set up development environments and Git/GitHub workflows, improving collaboration and reducing onboarding time by 30–40%. • Built frontend using Django templates/HTML, managed SSL certification and deployment, and drove adoption through documentation and onboarding of internal teams.

Intern at Danske IT and Support Services India Pvt Ltd, Jan 2022 - Jul 2022 · 7 mos, Bengaluru, Karnataka, India

Education

Bachelor of Technology, Computer Science and Engineering, Vellore Institute of Technology

Delhi Public School - R. K. Puram