AETHER MIND

Aether Mind — Wrocław & Hyderabad

We build AI systems which write the books!

Aether Mind is the company behind PenMyBook.ai, a production system that turns a topic into a finished, long-form technical book — researched, written, and bound end to end by a multi-agent AI pipeline.

This page exists for one reader: the Azure OpenAI partnerships team reviewing our request for GPT‑5.5 model access. It lays out what we've built, who's building it, and why this specific workload needs this specific model.

01 Product

PenMyBook.ai

A person gives PenMyBook.ai a topic, a few seed sources, and a chapter outline. From there, a pipeline of specialized agents takes over: one plans the book's structure, one researches each chapter against real sources, one writes the prose, and one reviews it — before the whole thing is bound into a finished, print- and Kindle-ready book.

It is live in production today, generating complete books for real users — not a demo, not a prototype pipeline run once for a screenshot.

plan → research → write → review → bind print PDF · KDP interior · Kindle package live in production
Generated book cover: an urban-planning and systems title
Cover · Systems & Cities
Generated book cover: a PostgreSQL internals title
Cover · PostgreSQL Internals
Generated book cover: a Redis internals title
Cover · Redis Internals
A generated chapter page, shown in paginated book reading mode
Interior · Reading mode
Generated book cover: a data engineering title
Cover · Data Engineering
Generated book cover: a Python data structures and algorithms title
Cover · Python DSA
A generated technical diagram explaining the cache-aside pattern
Diagram · Cache‑aside pattern
Generated book cover: an AWS SQS title
Cover · Queues on AWS
Generated book cover: an AWS EventBridge title
Cover · Event‑Driven AWS
Generated book cover: a probability theory title
Cover · Probability
A generated chapter page from the probability book, shown in paginated book reading mode
Interior · Reading mode
02 Track Record

Eighteen years of enterprise delivery, applied to AI

Aether Mind's founder, Appurv Saini, spent his career leading supply-chain and warehouse-execution programs for global manufacturers — the kind of work where a botched go-live stops a production line. That same delivery discipline — careful scoping, phased rollout, hypercare after launch — is what governs how we ship AI systems.

  • Pharmaceutical manufacturer (US rollout)S/4HANA embedded EWM
  • CanpackDecentralized EWM migration
  • Arla FoodsLabor management & forecasting
  • Ontex GlobalGlobal EWM template rollout
  • McCormickWarehouse automation & MES integration
  • EY Global Delivery ServicesDecentralized EWM, technical lead
  • 3MEWM/TM/GTS, IT team manager
  • Deloitte · HCL · DuPont · DellSAP SCM & supply chain delivery
SAP SCM EWM 9.0 Certified SAP C_S4EWM_2020 Certified SAP Activate Project Manager Certified
03 Why GPT‑5.5

The ask, plainly

Generating a full-length technical book isn't a single prompt. It's dozens of chapters, each reasoning over a large body of retrieved research, holding a consistent voice and continuity across tens of thousands of words, and coordinating multiple agents across planning, research, writing, and review. We currently run this pipeline in production on GPT‑5.4 via Azure OpenAI Service.

GPT‑5.5's longer effective context and steadier long-horizon reasoning are a direct fit for this workload, not a nice-to-have.

Higher quality, fewer retries

Better instruction-following across a 40+ chapter generation run means less rewriting and fewer flagged sections per book.

Lower latency at our scale

We operate a queue-based generation pipeline today; elevated quota reduces per-chapter wait time without changing that architecture.

Headroom to grow

Book and chapter volume can scale without degrading the quality bar we hold today.

Not a cold start

This is quota for a workload already running in production, with an existing Azure OpenAI integration.

04 Leadership

Two-person leadership, complementary disciplines

CEO & Founder

Appurv Saini

18+ years leading enterprise SAP supply-chain and warehouse-execution programs for global manufacturers, including Arla Foods, 3M, DuPont, and McCormick. Applies that same delivery discipline — scoping, phased rollout, hypercare — to shipping a production AI system.

LinkedIn ↗

CTO

Abhijeet Mishra

Senior Applied AI Engineer & Technical Architect. Ten years in software engineering, the last three focused on applied AI — building production agentic systems and LLM platforms, and training transformer models from scratch. Designs and operates the multi-service architecture behind PenMyBook.ai.

LinkedIn ↗
05 Tech & Trust

Built and run on production infrastructure

PenMyBook.ai runs as a set of independently deployed services on AWS, behind managed load balancing and TLS, with a queue-based generation pipeline that isolates long-running AI work from the user-facing API. This is live infrastructure serving real traffic today, not a local demo spun up for this request.

AWS-native, multi-service managed TLS & load balancing queue-isolated generation workloads
06 Contact

Get in touch

For questions about this request or about Aether Mind, reach us directly.