Saurabh Vishwakarma
Software Engineer / Backend and Systems
I design and build backend services, data pipelines, and cloud infrastructure. I write clean code, measure production latency, and keep systems reliable under high traffic.
“We have two lives, and the second begins when we realize we only have one.”
About
I engineer backend systems, data pipelines, and cloud infrastructure. I focus on high throughput, low latency, and operational simplicity.
Currently at Synthlane, I design and run core services in FastAPI and Go. I manage user authorization models, automate AWS deployments with Terraform, and monitor production tracing with Langfuse and Grafana. I care deeply about how systems behave under real load.
I enjoy diagnosing performance bottlenecks and eliminating unnecessary overhead. In production, I optimized database queries and service bottlenecks to cut API response times from 450 milliseconds down to sub 100 milliseconds at scale. I also built data pipelines using dbt and migrated core analytics infrastructure.
When I am not working on production systems, I study distributed systems internals, solve algorithmic challenges, and build low level tools from scratch, such as a concurrent Redis server written in Go.
B.Tech in Computer Science Engineering
RGPV University · Core focus: Distributed Systems, Operating Systems, Networks, and Databases.
Saurabh Vishwakarma
Backend Software Engineer
Experience
Synthlane
Software Engineer · Gurugram, Haryana
- Built modular FastAPI and Go backend services with granular role based access control for operators and workers.
- Cut production API latency from 450 milliseconds to under 100 milliseconds at 80 requests per second.
- Orchestrated database migration from PostgreSQL to Supabase with Fivetran and dbt for analytics.
- Automated AWS container provisioning and CI/CD releases using Terraform and GitHub Actions.
- Configured Okta and Supabase auth workflows with Langfuse tracing across distributed microservices.
Outlier AI
Software Engineer (Freelance) · Remote
- Architected a Streamlit evaluation platform to benchmark quality and latency across 20 machine learning models.
- Increased code generation accuracy by 15 percent using structured evaluation benchmarks and curated edge test cases.
- Validated model output safety, sandbox execution security, and automated regression suites.
Hunkaar Labs
Software Engineer Intern · Bhopal, India
- Deployed asynchronous task queues with FastAPI and Celery supporting over 5000 active users with 99.4 percent uptime.
- Integrated Razorpay payment workflows handling 700 transactions with end to end webhook verification.
- Built real time chat using WebSockets and low latency video rooms using LiveKit.