I build production-grade RAG systems, computer vision pipelines, and AI agents — turning complex ML research into working software that ships.
services
Specialized in the intersection of AI research and production engineering.
Custom retrieval-augmented generation pipelines — semantic search, vector databases, context compression, and multi-modal retrieval over your proprietary data.
End-to-end CV pipelines from data collection to deployed inference — object detection, image classification, OCR, and real-time video processing.
Autonomous agent systems that reason, plan, and use tools — from single-agent workflows to multi-agent orchestration with reliable human-in-the-loop checkpoints.
Complex CRUD systems, real-time dashboards, and API platforms — 6+ years building scalable backends and polished frontends that handle serious production load.
how it works
A clear, async-friendly process — from first call to handoff, you always know what's happening and what's next.
30-min call to understand your problem, constraints, and success criteria.
I deliver a written architecture doc with diagrams before any code is written.
Weekly async updates, shared staging environment, and early feedback loops.
Code walkthrough, documentation, deployment guide, and knowledge transfer.
A month of included support after handoff — bugs fixed, questions answered.
lab & work
Real projects — some anonymized due to NDA. More case studies being published.
Multi-tenant RAG system over 500K+ proprietary documents. Hybrid search (BM25 + dense), context reranking, and streaming responses. Reduced support ticket volume by ~40%.
Computer vision quality control system for manufacturing. YOLOv11 fine-tuned on custom dataset, ONNX-optimized for edge inference at 60fps. Deployed to Raspberry Pi + Coral TPU.
Autonomous research agent that takes a topic, searches the web, reads papers, synthesizes findings and produces structured reports. Built with LangGraph multi-agent orchestration.
Full-stack CRUD platform with real-time analytics, RBAC, webhooks, and a REST + GraphQL API. Handles 100K+ daily active users across 200+ tenant organizations.
client feedback
Feedback from teams I've built AI systems for — across RAG, computer vision, and full-stack.
“Akash delivered a RAG pipeline that genuinely works in production — not just a demo. Hallucination rate dropped below 2% within the first month. The evaluation loop he set up means we can actually measure quality over time.”
Marcus T.
CTO · LegalTech SaaS, USA
“We needed a computer vision system deployed on edge hardware in a warehouse with no stable internet. Akash scoped it correctly from day one, handled the thermal throttling issue before it became a problem, and delivered two weeks early.”
Priya S.
Head of Engineering · Logistics Platform, UK
“The architecture document he delivered before writing a single line of code was more thorough than what most senior engineers produce. It caught three design issues upfront that would have cost us weeks to fix later.”
Ahmed K.
Founder · AI Startup, UAE
“Async-friendly, clear updates every week, and zero surprises at handoff. The 30-day support window after delivery is something I wish every freelancer offered — we found two edge cases and they were fixed within hours.”
Sofia R.
Product Manager · HealthTech Company, Brazil
Names and company details shared with permission. Some details anonymized per NDA.
live demo
Try a live RAG demo — bring your own OpenAI key. Your key stays in your browser and is never sent to any server.
Stored only in memory. Never sent to our servers.
Enter your OpenAI API key above to start chatting.
This demo uses the OpenAI API directly from your browser. Your API key is never logged or transmitted anywhere except OpenAI's servers.
about
6+ years turning AI research into production software.
I'm Akash — an AI/ML engineer and full-stack developer based in Ahmedabad, India. I specialize in taking bleeding-edge ML research and shipping it as reliable, scalable production systems. From RAG pipelines serving thousands of queries a day to real-time computer vision on embedded hardware, I've built across the full spectrum of AI applications. I work with startups, agencies, and mid-size teams worldwide — primarily remote, always async-friendly.
faq
Everything you'd want to know before reaching out — answered upfront.
Have a different question? Send me a message
contact
Available for freelance projects, consulting, and long-term collaborations. Based in India, working globally.
Email directly
akashkp.freelancer@gmail.comTypical response time: within 24 hours
IST (UTC+5:30) — async-friendly worldwide
Remote-first
Working with clients across US, EU, Middle East, and Southeast Asia.
community
Follow My Journey
I share AI/ML tutorials, dev walkthroughs, and project breakdowns.
YouTube
Akash Code Cafe
AI/ML tutorials, project builds, code walkthroughs
Instagram
@akash_code_cafe
Dev updates, ML experiments, behind the scenes