SJMR Consulting & Development

Founded a consulting firm dedicated to supporting clients with marketing and creative technology initiatives, including education, game development, virtual production, digital advertising, omni-channel marketing, virtual events and developer relations.

Clients include: Capgemini Invent, MagickML, Crucible, Georgia State University, Polygonal Mind, Union Avatars, CyberBrokers, Causality

Founder and CEO

GigLabs

Orchestrated both long and short-term strategies for enterprise clients in the sports and entertainment industries, seamlessly bridging the gap between web3 and brand marketing for both crypto-native consumers and those without exposure to digital assets.

Created approachable educational content on web3 topics, such as customer loyalty, community management, pricing, supply strategies, and on-chain analytics to improve comprehension for clients and prospects.

Executed and optimized paid marketing campaigns for search, display, and brand partnerships using data-driven strategies, while coordinating cross functional teams to achieve successful partnership outcomes.

Partner Success Lead

Xaxis

Bridged multiple media agencies’ offerings including programmatic, search, social, SEO, and influencer media buys for a combined client portfolio, enabling additional revenue streams through client channel expansion.

Acted as cross-functional contact for data strategy, inventory, and our proprietary AI-driven media optimization tech.

Created, maintained, and presented sales, training, and pricing collateral for omni-platform media investment strategies.

Programmatic Lead

Led client relationships and book of business for all west coast agency clients while leading the Los Angeles office.

Grew total office revenue by 20% and internal product revenue by 17% in 2018 for programmatic ad spend.

Developed additional revenue channels with independent client relationships, accounting for 18% of office revenue in 2018.

Director, West

Managed a national team of 6 analysts, working across all partner agencies for over 30 clients.

Lead the Client Insights and Data Strategy group for the analytics team of >20 analysts, acting as a consultant for CRM and bespoke first and third party data activation tests.

Facilitated the Programming Project Group, educating and empowering analysts to incorporate R and Python into their work.

Manager, Data Strategy and Client Insights

Implemented a range of attribution systems and media studies, including MTA (Abakus, Logistic Regression Analysis), panel-based (Millward Brown, Lightspeed GMI), live intercept (Nielsen, Safecount, Research Now), tune-in (SambaTV) and in-store sales lift (Mastercard), to prove effectiveness of media at achieving client goals.

Utilized R and Python as back-end information aggregators to show the effect of non-media based data points on performance through internal data visualization platform to redefine client KPIs.

Trained and mentored entry level analysts, associates and interns through industry and skill based development.

Senior Analyst, Insights and Analytics

Managed campaign pacing and performance insights for all west coast GroupM clients through the design and maintenance of analytics dashboards that ingested ad server, viewability and fraud verification, DSP and audience data.

Queried internal DMPs to analyze and optimize cross-buy media interaction (reach, frequency, overlap, synergies, audiences) and attribution (conversion analysis, optimal revenue generating user-flow, media path analysis)

Delivered presentations demonstrating the effectiveness and role of digital media among the larger marketing portfolio to both agency members and key client stakeholders

Analyst, Insights and Analytics

Want to query your own documents locally and offline using the latest open source LLMs?

This allows you to use external data sources to embed 'knowledge' that can be retrieved based on your prompt and then fed in as context for your LLM queries.

In here, I run through setting up Mervin Praison ✅'s version of a local RAG pipeline using Ollama (for Llama 3), LangChain and Gradio. Super simple, super easy to get up and running.

Give it a thumbs up if you find these quick tutorials of value, and I'll keep making them.

Code: https://lnkd.in/g9MkrAAQ
Original Post: https://lnkd.in/gB4VPfXU

Stu Richards

So I'm building something myself (finally)...

A Custom Knowledge Podcast Generator: Free, Local and Open Source.

On the backend, I'm using: 
- Trafilatura to scrape the knowledge in a digestible manner
- Ollama for text inference with Llama 3
- Coqui XTTSv2 for TTS

This agent prototype scrapes data from any URL, creates a podcast script and then generates audio via TTS. I was able to put this together using just Anthropic Claude to help me through the weeds.

Next up is a UI, some optimization, length adjustment capabilities and maybe TTS fine tuning.

Stay tuned for more progress and in the wise words of Mitko Vasilev: Make sure you own your AI. AI in the cloud is not aligned with you; it's aligned with the company that owns it.

With Udio and Suno launching their new generative audio models, it begs a few questions:

- What will audio inpainting look like?
- Will generations have a DAW to edit/manipulate sound/structure?
- Will stems be accessible?
- Is prompt audio engineering going to be a thing? ("wide stereo", "in the style of Pink Floyd", "(TR-808: 1.4)")
- Could you prompt melodies and harmonies?
- When will accessible LoRA training be available?
- Can we go as far as recreating something like an IPAdapter for individual songs/samples?

The future of music is being determined as we speak, and it's apparent that generations could soon be indistinguishable from human authored releases. Having a strong manipulation toolkit, however, will greatly enhance what artists can do with this technology.

Late last year I started creating a non-exhaustive list of the things I tried to stay on top of on a daily basis.

After getting into the space a few years back, the scope of being domain SME kept creeping. Apart from tracking daily updates to each of these things, the list of things to keep track of kept expanding.

And it was exhausting, sucking up both time and mental energy which I knew would be better focused elsewhere.

I'm certain that a lot of you know this feeling.

So I took a little break. Minimal Discord. Minimal X. Minimal LinkedIn.

Turns out that you can still remain an SME and stay on top of all these things without having to be constantly 'live' within the conversations or wired into the feed.

Why do I say all this? Because I feel a hell of a lot better after realizing this.

If you're feeling burnt out by the sheer volume and scope of innovation that you're faced with on a daily basis, slow down. 

Sure, read the headlines, tap into your sources, get the download, but don't feel the need to be at the bleeding edge of the innovator subset of the adoption cycle. For the most part, the insights you derive from the time invested are likely not going to provide the ROI for you personally.

Align your time with your personal priorities, even if it seems like you'll fall behind, because in all likelihood, you won't.

With the release of Meta's Llama 3 yesterday, I wanted to share how quickly you can use this and other open source models locally using Ollama.

Note to self - don't ask LLMs about themselves during demo videos.

In all seriousness, I love how accessible Ollama makes interacting with new LLMs locally for testing and development purposes.

Hope you get to test out Llama 3 this weekend!

Taplio

Stu Richards's Linkedin Analytics

Stu Richards

What is Stu talking about?

Stu Richards's Best Posts (last 30 days)

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Stu Richards's Linkedin Analytics

Stu Richards

What is Stu talking about?

Stu Richards's Best Posts .css-qqfgvy{font-size:var(--chakra-fontSizes-md);}(last 30 days)

Want to drive more opportunities from LinkedIn?

Stu Richards's Linkedin Analytics

Stu Richards's Best Posts (last 30 days)