Headline

Building DeepSolv | Learning to ship fast & reliably

About

I’m Vishal Vishwajeet, a Founding Backend & AI Engineer at Deepsolv, where I work across applied AI, backend systems, cloud infrastructure, and production operations. Over the past three years, I’ve helped take Brandy—an AI agent for end-to-end social media engagement—from early ideation to production and product-market-fit iterations. My work includes building Python/FastAPI backends, developing LangGraph-based agent workflows, integrating LLMs and external platforms, and designing guardrails to improve the reliability and safety of AI-generated responses. I also manage and optimize cloud-native infrastructure across Kubernetes, EKS, Docker, ECR, Prometheus, Grafana, and AWS. I’ve contributed to major infrastructure migrations, including moving the platform from GCP to AWS and later migrating workloads across AWS regions from the US to Europe. These projects involved application migration, deployment restructuring, monitoring, networking, data services, and production cutover planning. Some of the outcomes I’m proud of include reducing cloud infrastructure costs by approximately 40%, bringing deployment times down to nearly one minute, implementing end-to-end observability and alerting, and leading root-cause analysis for issues such as memory leaks, high database load, and production downtime. Beyond engineering, I work closely with customers, product teams, and external partners such as AWS and Meta to solve production issues, improve onboarding, and translate real-world requirements into reliable product features. Earlier in my career, I built and deployed full-stack applications using the MERN stack for freelance clients and at RFLabs. I also enjoy building developer tools and personal projects, particularly around AI, backend systems, Go, and command-line applications. I’m most interested in building reliable AI products, scalable backend systems, and cost-efficient infrastructure that can perform well in real production environments.

Location

Bengaluru

AI stack

claude code

langchain

langgraph

mcp

openai

Pinecone

Work experience

Backend & AI Engineer at Deepsolv, Jun 2024 - Present · 2 yrs 3 mos, On-site, Built Brandy, Deepsolv’s first product—an AI agent for end-to-end social media engagement—from early ideation through production and product-market-fit iterations. - Built and scaled the Python/FastAPI backend, including LangGraph-based agent workflows, LLM integrations, external platform integrations, and production APIs. - Designed LLM guardrails and response-validation pipelines to prevent hallucinated, malformed, or policy-violating outputs. - Managed AWS and Kubernetes infrastructure across EKS, ECR, load balancers, deployments, monitoring, and production operations. Also performed Inter-cloud platforms and Inter-region infrastructure migrations. - Reduced cloud infrastructure costs by approximately 40% and streamlined deployments to nearly one minute. - Implemented Prometheus/Grafana observability, custom application metrics, Slack alerts, and incident-response workflows. - Designed usage-based pricing, plan entitlements, user-credit accounting, and LLM token-cost mapping. - Led root-cause analysis for memory leaks, database load, downtime, and infrastructure bottlenecks. - Worked directly with customers and external teams, including AWS and Meta, to resolve issues, improve onboarding, and deliver product and infrastructure enhancements.

Full Stack Engineer at Deepsolv, Dec 2023 - Jun 2024 · 6 mo, Hybrid, Built a Recommender System w/ Deep Learning on an Experimental Project FeedSolv that suggests Content for Linkedin Creators based on their Profile. Also, worked on building an Autonomous Voice Calling Agents (Archived), where I got to know a bit about some technologies like Redis, K8s for managing the deployments, and a few more. Learnings - - Coding in an organisation, requires really well structured+documented code. Learned about multi-layered Application from having layers for DB interaction uptill the Service layer to expose methods/routes. - Hands on, on creating basic Recommendation System using TFiDF, Word2Vec, Bag of Words etc Algorithms. - Handling Deployments on Kubernetes (GKE), for a well planned load balancing & minitoring systems via Prometheus metrics.

Fullstack Web Developer at Arthlex Research Pvt. Ltd., May 2023 - Aug 2023 · 4 mos, Chennai, Tamil Nadu, India, Developed the Web Interface for VIKRAM AI Bots, an initiative by ARTHLEX based on NextJs and Flask Backend.

Full Stack Developer at RFLABS, Apr 2023 - Nov 2023 · 8 mos, Chennai, Tamil Nadu, India, Led several Projects of company's clients which included Web & Mobile Apps. Helped getting the Projects completed before deadlines, maintained an internal Project Tracking Application. Built Projects using various Tech Stacks (MERN, Flask, Next.js) including some DevOps skills (Docker images, AWS EC2 & Azure Linux Deployments)

Education

Bachelor of Technology - BTech, Electronics and Communications Engineering, Delhi Technological University (Formerly DCE)

12th, PCM with Computer Science, kendriya vidyalaya

Public links

github: https://github.com/vishal-1230

medium: https://medium.com/@vishalvishwajeet

leetcode: https://leetcode.com/vishalvishwajeet422