Career
Experience
Laiqa (Startup)
Overview
After leaving Adani, I joined Laiqa as Co-Lead of Technology, a health-tech startup focused on female health. When I joined, the infrastructure was chaotic — high AWS bills, scattered codebases across multiple backend services in PHP, JavaScript, and Python, separate frontend applications, internal tools, multiple SQL and NoSQL databases, and disorganized AWS infrastructure with no proper environment separation. In 3 months, we rebuilt everything from scratch — unified the entire stack with PostgreSQL RDS, implemented proper dev/stage/prod environments, migrated all databases, deployed the new system live, and reduced AWS costs by 75%. Since then, we have pivoted to building Anya — an AI healthcare companion for women's health, specifically targeting PCOS/PCOD. As the sole engineer in the company, I am building the entire engineering stack from the ground up.
My Role
Technology Lead
Day-to-Day
As the only engineer in the company, I own the entire technology roadmap. My days are split between building Anya — a WhatsApp-first AI health companion — and making Laiqa an AI-native company. I built Anya's backend on Google ADK TypeScript with Claude Opus 4.7, managing WhatsApp Business Platform integration end-to-end. I am currently creating a custom dataset to fine-tune a LoRA on Gemma 4B for our in-house model, working closely with our doctors and dieticians to encode their clinical knowledge into the training data. I handle all bot deployments across stage and production environments, build and maintain internal tools for the team, and have deployed my own AI agents that run 24/7 for monitoring logs, tracking deployments, and driving continuous improvements. My deployed agents autonomously scan team conversations on Slack and user feedback, creating GitHub issues on their own. I run daily sweeps in Slack for the whole team and continuously push to move faster. Every system I build reinforces the human-in-the-loop paradigm, ensuring our doctors and dieticians validate AI outputs before they reach users. I also work across the Lang ecosystem (LangChain, LangGraph, LangSmith, Langfuse), with OpenRouter as my model gateway — constantly evaluating and testing the latest models as they ship.
What I Learned
- Pivoted the company from a generic health platform to an AI-first healthcare companion targeting PCOS/PCOD.
- Built Anya, a WhatsApp-first AI health bot, from scratch using Google ADK TypeScript with Claude Opus 4.7.
- Sole engineer handling the entire engineering stack — backend, deployments, internal tools, monitoring, and CI/CD.
- Deployed 24/7 AI agents for automated log monitoring, deployment tracking, and continuous system improvements.
- Creating a custom training dataset with in-house doctors and dieticians to fine-tune a LoRA on Gemma 4B for our in-house model.
- Integrated WhatsApp Business Platform end-to-end for patient-facing AI interactions.
- Built human-in-the-loop validation pipelines ensuring doctors review AI-generated recommendations before delivery.
- Deployed autonomous agents that scan Slack conversations and user feedback to auto-create GitHub issues, keeping the team aligned and moving fast.
- Reduced AWS infrastructure costs by 75% through complete infrastructure redesign and optimization.
- Work across the Lang ecosystem (LangChain, LangGraph, LangSmith, Langfuse) with OpenRouter as a model gateway for continuously testing new models, and fine-tuning small models locally and on rented GPUs.
Laiqa (Startup)
Overview
After leaving Adani, I joined Laiqa as Co-Lead of Technology, a health-tech startup focused on female health. When I joined, the infrastructure was chaotic — high AWS bills, scattered codebases across multiple backend services in PHP, JavaScript, and Python, separate frontend applications, internal tools, multiple SQL and NoSQL databases, and disorganized AWS infrastructure with no proper environment separation. In 3 months, we rebuilt everything from scratch — unified the entire stack with PostgreSQL RDS, implemented proper dev/stage/prod environments, migrated all databases, deployed the new system live, and reduced AWS costs by 75%. Since then, we have pivoted to building Anya — an AI healthcare companion for women's health, specifically targeting PCOS/PCOD. As the sole engineer in the company, I am building the entire engineering stack from the ground up.
My Role
Technology Lead
Day-to-Day
As the only engineer in the company, I own the entire technology roadmap. My days are split between building Anya — a WhatsApp-first AI health companion — and making Laiqa an AI-native company. I built Anya's backend on Google ADK TypeScript with Claude Opus 4.7, managing WhatsApp Business Platform integration end-to-end. I am currently creating a custom dataset to fine-tune a LoRA on Gemma 4B for our in-house model, working closely with our doctors and dieticians to encode their clinical knowledge into the training data. I handle all bot deployments across stage and production environments, build and maintain internal tools for the team, and have deployed my own AI agents that run 24/7 for monitoring logs, tracking deployments, and driving continuous improvements. My deployed agents autonomously scan team conversations on Slack and user feedback, creating GitHub issues on their own. I run daily sweeps in Slack for the whole team and continuously push to move faster. Every system I build reinforces the human-in-the-loop paradigm, ensuring our doctors and dieticians validate AI outputs before they reach users. I also work across the Lang ecosystem (LangChain, LangGraph, LangSmith, Langfuse), with OpenRouter as my model gateway — constantly evaluating and testing the latest models as they ship.
What I Learned
- Pivoted the company from a generic health platform to an AI-first healthcare companion targeting PCOS/PCOD.
- Built Anya, a WhatsApp-first AI health bot, from scratch using Google ADK TypeScript with Claude Opus 4.7.
- Sole engineer handling the entire engineering stack — backend, deployments, internal tools, monitoring, and CI/CD.
- Deployed 24/7 AI agents for automated log monitoring, deployment tracking, and continuous system improvements.
- Creating a custom training dataset with in-house doctors and dieticians to fine-tune a LoRA on Gemma 4B for our in-house model.
- Integrated WhatsApp Business Platform end-to-end for patient-facing AI interactions.
- Built human-in-the-loop validation pipelines ensuring doctors review AI-generated recommendations before delivery.
- Deployed autonomous agents that scan Slack conversations and user feedback to auto-create GitHub issues, keeping the team aligned and moving fast.
- Reduced AWS infrastructure costs by 75% through complete infrastructure redesign and optimization.
- Work across the Lang ecosystem (LangChain, LangGraph, LangSmith, Langfuse) with OpenRouter as a model gateway for continuously testing new models, and fine-tuning small models locally and on rented GPUs.