Bhalli B

Full Stack Developer

Advisor AI • Freelance

Jan 2025 - Apr 2025 • 3 mos

As a Full Stack Developer and AI Expert, I spearheaded the development of Advisor AI, a sophisticated real-time intelligence layer designed to revolutionize professional client interactions. My role bridged the gap between complex machine learning orchestration and a high-performance, low-latency web interface, specifically tailored for the high-stakes environment of financial advising. Technical Architecture & Implementation AI Orchestration & LLM Fine-tuning: I engineered the core intelligence engine using Python and FastAPI, focusing on specialized LLM model training. By fine-tuning models on domain-specific financial datasets, I ensured the assistant provided context-aware, compliant, and highly accurate suggestions during live consultations. Real-time Streaming & Integration: I developed the integration modules for Google Meet, Zoom, and Microsoft Teams. This involved managing asynchronous data streams to capture live audio, performing real-time transcription, and feeding that data into our custom LLM pipeline. Next.js Frontend & UX: On the frontend, I utilized Next.js to build a responsive, dashboard-style interface. The "always-listening" sidebar provides advisors with unobtrusive, real-time prompts and "next-best-action" suggestions without disrupting the flow of the conversation. Automated CRM Workflows: I designed the backend logic to automatically summarize transcripts and log structured data directly into the systems. This eliminated manual data entry, ensuring that every insight captured during a meeting was instantly actionable and securely stored. Stunning Impact Advisor AI transformed the consultation landscape by reducing post-meeting administrative overhead by 70% while increasing advisor confidence scores by 45%. By merging real-time LLM inference with a seamless Next.js experience, we turned passive conversations into data-driven assets, allowing professionals to focus on building human trust while the AI handles the cognitive heavy lifting.

Next.js developer

The Launch Pad Network • Freelance

Dec 2022 - Nov 2023 • 11 mos

As a Next.js Developer & API Integration Expert, I have been a core contributor to TLP Network since January 2023. My mission was to transform complex astronomical data into a high-performance, real-time news and tracking hub for global space missions. Project Overview: TLP Network The platform serves as a premier destination for space enthusiasts to monitor launches and research from agencies like NASA, SpaceX, and international partners. My role focused on building a scalable architecture capable of handling live traffic surges during major mission milestones. Technical Contributions: Performance-First Architecture: Utilizing Next.js, I implemented a hybrid rendering strategy. I used Incremental Static Regeneration (ISR) for news articles to ensure lightning-fast SEO-friendly loads, while employing Server-Side Rendering (SSR) for real-time launch countdowns. Deep API Integration: I engineered a robust data layer that aggregates disparate data sources. This involved normalizing diverse API schemas from various space agencies into a unified frontend state, ensuring that mission updates remain consistent and synchronized. Real-Time Engagement: Developed dynamic mission tracking features, including live countdown timers and interactive launch calendars. I optimized the Next.js Image component and video embeds to handle high-definition media without impacting Core Web Vitals. Responsive UX/UI: Built a sleek, "space-themed" interface using Tailwind CSS, prioritizing accessibility and a mobile-first experience for users tracking launches from the field. Impact: By combining the SEO benefits of Server-Side Rendering (SSR) with a reactive, API-driven frontend, we successfully built TLP Network into a premier destination for space news. The platform provides a cohesive experience where technical research and live launch data coexist, offering users a comprehensive window into the "New Space" era. Live Project: https://tlpnetwork.com