Co-founder, Bits&Bytes · Technical Lead · a3ro.dev
Co-founder of Bits&Bytes.
I build AI products
for the real world.
Technical lead, research engineer, and community builder. I design retrieval, automation, and AI-native systems where reliability matters as much as speed.
02 — Point of view
I care about the edge cases where systems usually break.
Most of my work comes from asking what happens when the system fails — not just when it succeeds. That is why I keep returning to model efficiency, retrieval quality, and software that has to survive real constraints instead of performing well in ideal demos.
The same instinct drives the community work. Bits&Bytes and STEMist exist because the gap between ambition and a first build should be smaller than it is.
03 — Trajectory
Work that compounds across product, AI, and community.
I like work where technical execution and systems thinking overlap: shipping product, building AI-native workflows, and creating ecosystems that make more builders possible.
Co-founder & Technical Lead at Bits&Bytes
Leading product, engineering, and digital infrastructure for gobitsnbytes.org while shaping Bits&Bytes into a high-signal home for builders, workshops, launches, and community programs.
STEMist Chapter Lead at STEMist Education
Founding the club at Boys High School, Prayagraj — zero gate-keeping, builder-focused STEM with hands-on workshops, demo days, and practical engineering culture.
Jr. Research Engineer at jhana.ai
Built AI-powered legal tech systems — conversational flows for WhatsApp using LLM voice transcription with Gemini, natural language command parsing for automated document editing, concurrent processing for real-time message streams at scale, and voice-to-text pipelines for hands-free legal document workflows.
Blockchain / HyperLedger Intern at Shihaan Tech Global
Contributed to team projects, updated and improved the company website, and built internal tools to streamline operations using TypeScript, Python, FastAPI, and LangChain.
04 — Selected builds
Projects shaped by reliability, systems thinking, and real-world usefulness.
I am most interested in products where the technical architecture changes what the experience can actually do, not just how it looks in a launch post.
ARIA Autonomous Agent
Autonomous execution stack for long-running tasks, memory-aware planning, and reliable tool orchestration across complex workflows.
Built for complex execution where observability and steady planning matter more than theatrical autonomy.
MedReady AI
Workforce-readiness platform helping aspiring healthcare professionals train, assess, and move faster in underserved regions.
Focused on access, clarity, and a product flow that keeps high-friction users moving.
HybridMind
Retrieval architecture that fuses vector semantics with graph structure for sharper reasoning, recall, and context quality.
Explores how smaller or constrained models can recover capability through better external knowledge systems.
Lexana
Legal intelligence assistant tuned for Indian workflows, grounded research, and long-context retrieval that stays useful under pressure.
Designed around the simple test of whether it makes real legal work faster, cleaner, and easier to trust.
Bits&Bytes Platform
Digital home for gobitsnbytes.org—community, programs, launches, and member experiences brought into one cohesive system.
Turns community operations into durable infrastructure instead of scattered tools and manual coordination.
Cognitive Note Architecture
Private desktop knowledge workspace for encrypted note-taking, retrieval, and local-first AI assistance.
Built for people who want useful intelligence without giving up privacy or local control.
Educational Resource Matrix
High-signal education platform that organizes quality learning resources into a clean, searchable system people actually want to use.
A systems answer to information overload: better structure, better discovery, less wasted effort.
05 — Stack & interests
The tools are flexible. The taste is consistent.
I move across product surfaces, AI workflows, infrastructure, and education systems. The stack changes. The goal stays the same: build something clear, dependable, and actually useful.
Core stack
What I keep returning to
05 — Contact
Need someone who can think through the system and still ship the product?
I am most useful on work that has real constraints: AI product flows, retrieval systems, internal tools, community platforms, and software that has to stay clear under pressure.
Book a callBest for
AI systems, retrieval, product engineering
Usually reply
Within 24–48 hours
akshatsingh14372@outlook.com
Schedule
calendly.com/a3ro-dev