Private AI products. Production systems. Real-world impact.
I build AI that people can actually ship and trust — from PocketMind Hybrid AI (offline-first desktop & mobile AI for Mac, Windows, Linux, and Android) to CodexBridge and production systems at Zindex. BS AI @ NUST, Islamabad. Topped ETEA ’25 CS (1st in KPK Computer Science).
About Me
I'm Muhammad Nouman Shakeel — Founder of PocketMind, Co-founder of CodexBridge, AI Engineer at Zindex, and a BS Artificial Intelligence student at NUST, Islamabad.
I build AI systems that solve real problems — not concepts, not notebooks, but production-ready products people can actually use. PocketMind Hybrid AI is my flagship: private offline AI by default, online power when you need it — for Mac, Windows, Linux, and Android.
At CodexBridge, I'm closing the gap between what a product does and what teams actually understand as they ship faster. At Zindex, I build production-style AI systems in parallel with founding work.
Beyond products, I topped ETEA ’25 CS (1st in KPK Computer Science), topped Pakistan twice in English Language Proficiency Olympiads, represented the country internationally in Turkey, and was named Student of the Year at The Peace College.
Products & Focus
Core AI / ML
Engineering
Career
Flagship Product
Private Offline AI. Online power when you need it.
Your data stays yours. Offline by default. Online when it matters.
✦ Chat, files, images, voice, custom AI, and Knowledge Chat — one app for Mac, Windows, and Linux, with Android support for portable private AI.
PocketMind Hybrid AI is a desktop-first AI workspace (Mac · Windows · Linux) with a polished Android companion. Run models locally with llama.cpp, attach documents, index folders with Knowledge Chat (RAG), analyze images, talk with voice, use GPU acceleration (Metal / CUDA / Vulkan), and connect online models only when you choose. Built for founders, analysts, and teams who need real capability without giving up control of their data.
Run models on your machine with llama.cpp. Prompts stay local when you choose offline mode — built for compliance, IP protection, and sensitive work.
Connect Gemini, Groq, Hugging Face, Together AI, and more when you need cloud speed or vision — only when you opt in.
Index a folder and ask grounded questions with sources. Answers from your files — not guesswork from the open internet.
Attach PDFs, DOCX, text/code, CSV, JSON, logs, and work through documents in one clean workflow.
Understand screenshots, diagrams, and visuals. Generate images offline or via optional online providers.
Talk instead of only typing — voice input, replies, and stop controls for natural hands-free use.
Personas and system prompts tuned to your workflow — assistants that match how you actually work.
Triage, investigation support, reports, and company-knowledge search built for security teams.
Point clients at your company’s private OpenAI-compatible endpoint — AI inside the company boundary.
Metal / CUDA / Vulkan with smart CPU fallback, plus diagnostics and recovery tools for real deployment.
Backup & restore, storage controls, and security settings designed for real machines — not demos.
Offline by default. Online only when it matters. No developer cloud collecting your private conversations.
Portfolio
Privacy & Trust
PocketMind’s full privacy policy is available inside this same website as a dedicated hidden section. It stays out of the normal scrolling experience, but opens instantly when needed for users or Play Store review via #privacy-policy.
Open Full Privacy Policy →Full Policy
This full policy is intentionally hidden from normal page scrolling. It opens only through the Privacy button or a direct Play Store link ending in #privacy-policy.
Back to Privacy Summary →This Privacy Policy describes how PocketMind Hybrid AI for Android (“PocketMind,” “the App,” “we,” “our,” or “us”) handles information when you install and use the App. PocketMind is a hybrid AI assistant that can run AI features on your device (offline) and, when you choose, connect to third-party online AI services, your organization’s private AI server, Google Play Billing, and related platform services.
By downloading, installing, or using the App, you acknowledge that you have read this Privacy Policy. If you do not agree, do not use the App’s online, voice, image, organization-server, or paid features that require data processing beyond local on-device storage.
1.1 What this policy covers. This policy covers the PocketMind Hybrid AI Android application distributed through Google Play (application identity associated with PocketMind Hybrid AI) and related in-app features described below.
1.2 What this policy does not cover. This policy does not govern websites, APIs, models, or servers operated by third parties (for example Groq, Google Gemini / Google AI, OpenRouter, Hugging Face, Together AI, Pollinations, OpenAI, ElevenLabs, your employer’s organization server, or Google Play). When content leaves your device to those services, their privacy policies and terms apply in addition to this policy.
1.3 Controller / developer. For local app storage and the PocketMind product experience, the developer is Muhammad Nouman Shakeel (PocketMind), Pakistan. For purchase processing, Google LLC / Google Play acts as payment processor. For AI inference you request online, the selected provider acts as an independent processor/controller of that request content under its own terms.
PocketMind does not currently operate a developer-owned backend that creates user accounts for PocketMind, hosts your chat history in a PocketMind cloud inbox, or sells your conversation content to advertisers.
In particular, we do not intentionally:
Important limitation: “We do not collect on PocketMind servers” does not mean “nothing ever leaves your phone.” Content leaves your phone when you use online AI, downloads, organization servers, or Google Play. Android, Google Play Services, and network carriers may also process technical data according to their own policies.
Depending on how you use the App, PocketMind may store information in on-device app storage (and related Android storage you authorize), including without limitation:
Local storage remains on your device until you delete it, clear app data, uninstall the App, or otherwise remove those files. Device backups (Android backup / manufacturer cloud backup), if enabled by you or your device settings, may copy app data off-device under those systems’ rules.
PocketMind offers a freemium experience. Exact limits may change with product updates, but as of this policy’s last update the App is designed around the following model:
pocketmind_pro_monthly and pocketmind_pro_yearly). Eligible upgrades from a prior paid app version may receive a lifetime unlock recorded locally and validated against Play where applicable.Entitlement state may be stored on-device and refreshed through Google Play Billing purchase/restore flows. We do not invent a separate PocketMind payment account; billing identity is handled by Google under your Google account.
PocketMind allows you to enter API keys / tokens for supported providers. Those credentials are stored locally on your device and are transmitted from your device to the selected provider when you validate a key or run a request.
You are solely responsible for:
When you select offline/local models, PocketMind processes prompts and generates responses on your device using on-device model runtimes (for example local LLM inference). In true offline mode, PocketMind does not send those prompts to online chat providers.
Offline features may still require Internet for one-time actions you initiate, such as downloading a model file from a hosting provider (for example Hugging Face model repositories). After download, local inference can proceed without sending chat content to that host.
Offline performance and availability depend on your hardware (CPU/GPU/Vulkan support where used), RAM, storage, thermal limits, and selected model size. Some catalogued neural voice engines may be listed for future/on-device use while the currently guaranteed offline speech path uses Android System Text-to-Speech where implemented.
If you select an online chat model, PocketMind sends request data from your device directly to the selected provider over the Internet. Depending on the model and settings, this may include:
PocketMind does not control how third-party providers log, retain, train on, review, or further process content after receipt. Review each provider’s privacy policy before sending personal, confidential, biometric, medical, legal, or otherwise sensitive information.
If you configure Organization Server mode, PocketMind connects to a server address and access token that you (or your organization) provide. Expected integrations are OpenAI-compatible private gateways (for example vLLM, llama.cpp server, LiteLLM, LocalAI-style proxies, or a company-hosted gateway).
In this mode, prompts, context, and related request data are sent to that server under your organization’s control and policies. PocketMind is not responsible for your organization’s retention, monitoring, HR, security, or compliance practices. Free-tier online quota accounting in the App may treat organization-server usage as counting toward free online usage limits where the product logic so provides.
You may attach or import files such as PDF, DOCX, TXT, code, CSV, JSON, logs, and similar formats. The App may extract text or otherwise prepare content for the selected AI mode.
Do not attach files you are not authorized to process. You are responsible for complying with workplace, school, and confidentiality rules.
For vision-capable online models, attached images may be encoded and transmitted to the selected provider. Offline text-only models typically cannot interpret raw image pixels. Image contents may reveal faces, locations, documents, or other sensitive information—review before sending online.
Image Studio lets you generate images offline (local image models when available) or online (providers such as Hugging Face, Together AI, Pollinations, and other catalogued image endpoints). Access to Image Studio is a PocketMind Pro feature.
When you generate online, your prompt, style/size/detail settings, and authentication material may be sent to the selected image provider. Generated images may be saved locally if you choose. Offline generation keeps the prompt/model work on-device for that path, aside from any model download you previously initiated.
Voice Studio provides text-to-speech generation and voice-cloning workflows. Access requires PocketMind Pro. Depending on the selected tier and model:
Microphone and audio. The App may request microphone / audio recording permission (Android RECORD_AUDIO) so you can capture speech or reference clips. Audio may also be imported from files you select. Synthesized outputs may be written to app-private storage (for example a voice output directory) and played back on device.
High-risk disclosure — voice cloning: Reference audio and cloned-voice outputs can be biometric or otherwise sensitive. If you upload another person’s voice, you must have all rights and consents required by law. Do not use Voice Studio to impersonate someone without authorization, to commit fraud, or to violate platform or provider rules. Online cloning transmits reference audio to the selected provider; PocketMind cannot delete data from that provider on your behalf.
Subscriptions and in-app purchases are processed by Google Play Billing. Google may receive and process purchase tokens, product identifiers (for example monthly/yearly Pro SKUs), order data, and account/device information necessary to complete, acknowledge, restore, or refund purchases under Google’s Privacy Policy and Google Play Terms.
PocketMind may store local entitlement flags so the App can unlock Pro features. Purchase disputes, refunds, and payment methods are handled through Google Play, not through a PocketMind payment portal.
When you download offline chat or image models, the App may contact hosting platforms (commonly Hugging Face repositories or similar URLs in the catalog) to fetch model files. Those platforms may log IP address, request metadata, and authentication if a token is required. Model files can be large and remain on your device until deleted.
Depending on Android version and the feature you use, PocketMind may request or use:
You can deny non-essential permissions; related features may then be unavailable.
Depending on your choices, PocketMind may communicate with services such as:
Provider catalogs change over time. The in-app Models / Voice / Image hubs are the authoritative list of currently offered integrations for your installed version.
PocketMind is designed around AI product functionality. The App does not currently integrate a separate third-party advertising SDK for the purpose of selling your chat content to advertisers. We also do not currently operate PocketMind-owned product-analytics servers that ingest your chat transcripts.
Nevertheless, Android, Google Play, crash/diagnostics facilities you enable at the OS level, and third-party AI providers may collect technical or request logs independently. Those activities are governed by those parties’ policies.
We do not sell your personal data. Information may leave your device when:
On device: Local data remains until you delete chats/keys/models/outputs, clear app storage, or uninstall. Uninstalling removes app-private storage but does not erase data already sent to third parties, nor files you exported to shared storage.
With third parties: Retention is controlled by the provider or your organization. To delete provider-side data, use that provider’s tools or contact them directly. PocketMind cannot guarantee deletion from third-party systems.
Requests to the developer: Because we do not host your chat history on a PocketMind cloud account, we generally cannot retrieve or remotely wipe device-local chats. Contact us if you need help understanding deletion steps on your device.
We design the App to limit processing to what a feature needs and to keep keys and chats in app storage by default. No method of electronic storage or transmission is completely secure. You are responsible for device lock screens, OS updates, shared-device hygiene, exported backups, and API key rotation.
PocketMind is not directed to children under 13 (or the minimum age required in your jurisdiction for AI / online services without verifiable parental consent). We do not knowingly collect personal information from children under 13 on a PocketMind-operated server. If you believe a child has provided sensitive information through online providers via the App, stop use, delete local data, revoke API keys, and contact the relevant provider and us as appropriate.
Do not use PocketMind as a substitute for professional medical, legal, financial, mental-health, or other licensed advice. AI outputs may be wrong or unsafe. If you process special-category or regulated data (health, biometrics, government IDs, children’s data, trade secrets), you must ensure you have a lawful basis and that any third-party provider you select is appropriate for that data.
AI-generated text, images, and audio may be inaccurate, biased, infringing, offensive, or misleading. You are responsible for reviewing outputs before use. You agree not to use the App to violate law, infringe others’ rights, create non-consensual intimate imagery, commit fraud/impersonation, or breach provider acceptable-use policies.
If you are outside Pakistan (or outside a provider’s home region), online features may transfer your prompts, files, audio, images, and metadata to servers in other countries. Those countries may have different data-protection laws than your own.
Depending on your location, you may have rights to access, correct, delete, or restrict certain processing, or to lodge a complaint with a supervisory authority. Because most PocketMind content is stored on your device, many rights are exercised by using in-app deletion controls and Android settings. For Google Play purchases, use Google account/Play tools. For provider-held data, contact that provider. Contact us using the details below for questions about this policy.
We may update this Privacy Policy to reflect new features (including voice, image, freemium, billing, or provider changes), legal requirements, or security improvements. The updated version will be posted at this same public privacy policy location (this website’s #privacy-policy URL) with an updated “Last Updated” date. Continued use of the App after an update means you acknowledge the revised policy. Material changes may also be described in the Google Play listing where appropriate.
For privacy questions, deletion guidance, or support regarding this policy, contact the developer:
Muhammad Nouman Shakeel / PocketMind
Phone / WhatsApp: +92 340 5055603
Please do not send passwords or full API keys over unsecured channels. If you must discuss a compromised key, revoke it at the provider first.
Recognition
Credentials
Get In Touch
Have an AI app, chatbot, automation system, website, surveillance AI, or custom model idea? Fill the form and your project inquiry will open with a polished message already prepared from your details.