Anand Bhandari

Anand Bhandari

Principal AI Full-Stack Engineer · CTO @ SOLS Group

Building production AI systems — from an adaptive AI English-learning platform for Malaysian students, to an MCP server that lets AI assistants run company operations, to 105 shipped indie tools. I own AI architecture and hands-on delivery, and lead 20+ engineers.

14+ Years experience
105 Indie tools shipped
2 MCP servers built
20+ Engineers led

About

Principal-level full-stack engineer and CTO with 14+ years shipping web, mobile, and AI-powered products, now focused on building production LLM applications end to end. I own the architecture and hands-on delivery of AI features inside real products — an adaptive AI English-learning platform, an internal operations platform exposed to AI assistants through a permission-scoped Model Context Protocol (MCP) server, LLM pipelines that turn meeting transcripts into weekly reports, and an LLM-driven automated testing agent.

Multi-provider by design: Anthropic Claude, OpenAI, Google Gemini, OpenRouter, Groq, and local models via Ollama. Anthropic-certified across the Claude API, MCP, Agent Skills, and Claude Code, I run my daily engineering on an AI-native workflow I built myself. I lead 20+ engineers and ship AI features into production, not prototypes.

Areas of expertise: Production LLM Application Design · Multi-Provider AI Architecture · MCP Servers & Clients · AI Agents & Tool Use · RAG · Speech AI · Multimodal · Local / Privacy-First AI · Full-Stack Product Engineering · Technical Leadership

Experience

CTO / Principal Full-Stack Engineer & Solution Architect
SOLS Energy (SOLS Group), Malaysia
Sep 2022 – Present
  • Lead technology strategy, product architecture, and digital transformation across the group's web, mobile, and AI product lines; hands-on across the full stack while managing and mentoring 20+ engineers.
  • Architected and shipped ACE, an AI-powered English-learning platform for Malaysian SPM students — an adaptive learning engine that builds personalized study paths from diagnostic tests and gap analysis, with timed mock exams mirroring the national SPM format.
  • Built SOLS Boss, the group's internal operations platform (departments, tasks, work requests, SOPs, standups, meeting notes, reminders; HubSpot and Google Sheets sync), and exposed it to AI assistants through a permission-scoped MCP server over streamable HTTP — per-user hashed API keys with expiry, read / read-write grants with tool allow-lists and department scoping, rate limiting, and an audit log on every call.
  • Built an LLM weekly-reporting pipeline (Next.js app plus browser extension) that ingests meeting transcripts and generates structured weekly reports through a provider-agnostic, OpenAI-compatible LLM client with JSON-mode output, so the model and provider are switched at runtime, not in code.
  • Built an LLM-driven automated testing agent for ACE that runs smoke, guided, and exploratory test sessions from app maps and prompts, and writes reports.
  • Set the group's LLM-integration patterns: provider-agnostic clients across Anthropic Claude, OpenAI, Google Gemini, and OpenRouter, with Ollama for private on-premise workloads; chose models and providers against cost, latency, and accuracy.
  • Architected the renewable-energy web platforms and cloud infrastructure: mobile and PWA apps for real-time solar monitoring serving 10,000+ users, the solar-subscription self-care system, and the SOLS Green solar-financing journey.
  • Spearheaded the HubSpot API integration on Firebase Cloud Functions, and a CTOS credit-report integration for risk assessment; established CI/CD, code-review practice, and production-readiness gates across all squads.
Solution Architect | Web & Mobile Platform Innovator
SOLS Foundation (SOLS 24/7), Malaysia
Jan 2021 – Jan 2026
  • Oversaw all technical solutions and software process management for the non-profit.
  • Designed and maintained education platforms and internal systems serving 4,000+ families — Skills Hub, Solar Academy, and TVET Academy with an LMS for enrollment, tracking, and certification.
  • Led API integrations, the full software lifecycle (development, testing, deployment, maintenance), internal technical support, and mentoring of the development teams.
Mobile Application Developer / Team Lead
SOLS 24/7, Malaysia
Jan 2019 – Jan 2021
  • Managed Android, Web, and iOS projects and served as primary developer for the SOLS SMART, SOLS EDUCATION, and SOLS CLASSROOM apps and websites; built the cross-platform mobile apps in React Native from scratch.
  • Managed API integrations and backend systems; supervised testing, deployment, and agile processes.
  • Grew the engineering team from 5 to 20+ developers with CI/CD, code review, and quality gates.
Web & Mobile Developer
MINDIII Systems, India
2016 – 2017
  • Built Android and web apps, connected to back-end services via RESTful APIs and cloud messaging.
Senior Software Engineer
IBM, Bengaluru, India
Jan 2013 – Jul 2016
  • Built client-side web applications and mobile-responsive portal UIs within IBM global teams under Agile.
  • Served as Scrum Master and contributed to the IBM Technology Adaptive Program (TAP).
Software Developer
ManpowerGroup, India
2011 – 2012
  • Built websites and web applications using HTML5, CSS3, jQuery, WordPress, and Photoshop.
Web Developer
myMO IT Solutions, India
2009 – 2011
  • Built responsive websites and web applications using HTML, HTML5, CSS2/CSS3, jQuery, WordPress, and Photoshop.

Skills

AI & LLM
Anthropic Claude API, OpenAI API, Google Gemini, OpenRouter, Groq, Ollama & LM Studio (local LLMs, vision models), Model Context Protocol SDK (stdio & streamable HTTP), Agent Skills, Claude Code, AI agents & tool calling, LangChain, LlamaIndex, RAG, Embeddings & vector search (pgvector), Prompt engineering
Speech & Multimodal
Whisper / faster-whisper (STT), ElevenLabs, MiniMax, Piper, Edge TTS, Gemini & MiniMax image generation, Remotion & HyperFrames (programmatic video), vision models
Languages
TypeScript, JavaScript (ES6+)
Web & Front-End
React, Next.js (React Server Components, Server Actions), React Native, Expo (Expo Router, NativeWind), Qwik, Builder.io, Tailwind CSS, shadcn/ui, Radix UI, Redux, Zustand, TanStack Query, React Hook Form, Zod, TipTap, PWA
Mobile & Desktop
React Native, Expo, Android, iOS, Electron, Tauri
Back-End & APIs
Node.js, Bun, Express, Fastify, tRPC, Prisma, Drizzle ORM, Better Auth, pg-boss job queues, REST & GraphQL APIs, WebSockets, enterprise API design, system integrations
Cloud & DevOps
Google Cloud (GCP), Firebase, Cloud Functions, Vercel, Docker, CI/CD, GitHub Actions
Databases
PostgreSQL (+ pgvector), MongoDB, Firestore, Redis, SQLite, Supabase
Integrations
HubSpot API, CTOS, Mailgun, Stripe, Google Sheets, WordPress REST API, Respond.io, ClickUp, Linear
Leadership
Technical vision & roadmapping, team mentorship (scaled 5 → 20+ engineers), Agile product delivery, code reviews, stakeholder management, DevOps & deployment pipelines

Projects

AI Projects

NixBiz · Indie AI Lab
nixbiz.com
  • 105 tools shipped, open source on GitHub: 99 desktop apps (Windows & macOS), 2 Chrome extensions, 2 web apps.
  • AI subset: chatnix (private local chat and document Q&A on Ollama), asknix (drag a screen region, ask a local AI), meetscribe (meeting recorder, local transcription, AI summaries), subnix / scribenix (Whisper subtitles and transcripts), textgrab / pdfocr (screen and PDF OCR), bgone / upscalenix (on-device background removal and upscaling), PageEcho (Gemini page summaries read aloud), Smart Bookmarks (OpenAI / Gemini summaries and tags).
News-Tube · Automated AI News Channels
  • Automated TV-style news channels — typed segments self-render frames, tickers, anchor voiceover, and transitions; Gemini writes scripts and headlines, an Ollama vision model picks images, Remotion renders.
  • Ships its own stdio MCP server so coding agents can drop in content.
Screencast Creator · AI Video Tooling
  • Screen capture with hand-drawn annotation, keyframe motion, AI voiceover (ElevenLabs, MiniMax), and Groq Whisper word-level captions, exported via Remotion.
Milo Agent · Local-First Developer Workspace
  • Local-first workspace with terminals, prompt library, Monaco editor, agent runs and dispatch rules, shared memory, schedulers, and a milo CLI; Ollama chat with local tool calling.
Nix Translator · Mobile AI Translator
  • Expo React Native translator — OpenRouter as the LLM gateway, Gemini, Whisper speech-to-text, and ElevenLabs text-to-speech; API keys kept in secure storage.
Story-video Pipeline · AI Video Production
  • Narrated explainer and Shorts production from a script — MiniMax scene images, ElevenLabs voice with word timings, HyperFrames captions and render — driven by a guided Claude Code skill.

Platform & Product Work

SOLS Energy Platform · Renewable Energy
solsenergy.com
  • End-to-end solar platform (2,600+ installations, 30+ MW) — system design, installation booking, and real-time energy-monitoring PWA — integrating HubSpot CRM, payment gateways, and government incentive tracking into one lead-to-installation pipeline; plus the ChargeSingh EV-charging portal and Solar Academy.
SOLS Green · Green Fintech
sols.green
  • Green fintech with RM 0 upfront, instant solar financing, and transparent pricing; end-to-end journey from power study to installation with real-time status tracking; GA, Google Ads, and Meta Pixel conversion tracking.
SOLS Edu · E-Learning Platform
solsedu.com
  • Multi-language e-learning PWA (50K+ students, 4 languages) with live classes, self-paced courses, and a Cambridge English pathway.
SOLS Foundation · Non-Profit
sols.foundation
  • Digital ecosystem for Malaysia's tax-exempt non-profit, providing free education and employment support to the B40 population (4,000+ families served).
  • Delivered Skills Hub, Solar Academy, and TVET Academy platforms with LMS enrollment, tracking, and certification.

Open Source

WhisperDrop · Speech-to-Text
GitHub
  • Local hotkey speech-to-text, GPU-accelerated, no cloud required.
AI Text Rewriters · AI Text & TTS
GitHub
  • Multi-provider text rewriting via Groq, OpenRouter, Ollama, and Gemini, plus built-in Edge TTS.
FastSimple (Open Whisper) · AI Speech-to-Text
GitHub
  • Speech-to-text with Whisper, Ollama, and LanguageTool across five writing tones; auto-inserts text at the cursor via a global hotkey.
Piper TTS Web UI · Text-to-Speech
GitHub
  • Streamlit app for fast, local text-to-speech with multiple voices and per-sentence generation modes.

Education

Bachelor of Engineering
Rajiv Gandhi Technical University
2007 – 2011
  • Strong foundation in engineering principles, algorithms, data structures, and software design patterns.
  • 14+ years of continuous delivery, now focused on production AI systems and LLM application engineering.
  • Anthropic-certified across the Claude API, MCP, Agent Skills, and Claude Code.

Certifications

AI Fluency: Framework & Foundations Anthropic, 2026
Building with the Claude API Anthropic, 2026
Claude Code in Action Anthropic, 2026
Introduction to Agent Skills Anthropic, 2026
Introduction to Model Context Protocol (MCP) Anthropic, 2026
Introduction to Claude Cowork Anthropic, 2026
Generative AI: Prompt Engineering IBM, 2026
Basics of AI IBM, 2025
Generative AI: Introduction & Applications IBM
Introduction to AI IBM
Prompt Engineering Codecademy
Defending Against AI-Generated Attacks Codecademy
GitHub Copilot for Beginners: Write Software With AI Coursera Project Network
Introduction to Software Engineering IBM
Certified React Native Developer Udemy

Contact

Whether you need a technology leader, AI architecture consulting, or advisory for your team, I'm always open to meaningful conversations. Open to remote (global) and hybrid roles.