since february 2024, i've been working as an engineer and technical lead at avalon in berkeley, ca. alongside a talented team of three engineers, i've been developing a retrieval-augmented generation chatbot tailored for law firms. this chatbot automates eDiscovery and drafts motions and briefs by leveraging tens of thousands of case files using technologies like pinecone, nextjs, dynamodb, typescript, and go. i've also spearheaded the creation of a web socket file transfer pipeline to aws s3, ensuring efficient downstream feeding into our rag model. moreover, i've optimized our language models with langchain to assess the probability of case wins, aggregating metrics from various documents to enhance decision-making processes.

i was hired as a machine learning engineer at megalabs, an ai venture firm. i, along with four other engineers, created a react app using the t3 stack (typescript, trpc, tailwindcss) for a sports data conversational assistant. i also built an intelligent sql agent with aws bedrock, aws athena, langchain, gpt 4.0, and aws open search that could generate and run complex sql queries on its own to answer user questions in python. on top of that, i put together a mongodb-backed interface to keep tabs on thousands of user queries and check how accurate our models were during alpha testing.

since june 2023, i’ve been the go-to person for all things tech at opsm pro in ottawa, ca, as their lead full stack engineer. i built our main e-commerce site using the mern stack (mongo, expressjs, react, nodejs) from the ground up. integrating the stripe api, i set up a system to handle over a dozen products and subscriptions, helping us process more than $1,000,000 in payments each year. plus, i designed and implemented mysql databases to keep all our user and customer info safe.