DEV Community: AMELIA HOUGH-ROSS

Amazon Bedrock Beats Paper: Troubleshooting LLM Hallucinations

AMELIA HOUGH-ROSS — Sun, 08 Jun 2025 17:07:56 +0000

This post is meant to continue the story I started here.

Overview

As a reminder, I wanted to use LLM's to chat with my Jira ticket data, since that was all the hype in 2023-2024 thanks to the rise in popularity of chatgpt. I knew my ticket data contained specific use cases from my loudest customers because they called me on the phone, but I wanted to know about everyone else.

What questions were they asking?
What was my team missing?
What connections across the organization could be made to speed up research activities
How could my team create more helpful FAQ documentation?

I started this use case in January of 2024, using Claude V2.0 and Amazon Titan Text Embeddings and proceeded to spend the entirety of that year losing all faith that I could accurately chat with my data using Amazon Bedrock Knowledge Bases. I began to see improvement when Claude 3.5 came out, but not the robust results I expected. For more details on the use case, please check out this video from AWS re:Invent 2024 here.

While, I had a great plan, and solid execution, the results of that plan were not what I would have expected. By using the Amazon Bedrock Knowledge Base that originated solely as an OpenSearch Serverless Collection to create the vector store for my RAG pipeline, I continued to suffer through LLM hallucinations, and I had no faith that my queries were returning good responses. Here's how I knew:

Out of 4,000 Jira tickets, I asked my Amazon Bedrock Knowledge Base to categorize the Top 5 most requested items in Jira ticket data. It came back with 5 different things. One of which was Jupyter notebooks. I thought, Aha! How many tickets contain the spelling of Jupyter? I checked my CSV file and there were exactly 13. Thirteen out of four thousand. That's .3%. I wouldn't say that was a top 5 contender, but it gave me the detail I was looking for to test the RAG pipeline to confirm when I asked the knowledge base how many Jupyter tickets existed, that it gave me the correct response.

Awkward Data Responses

During 2024, the behavior I experienced showcased that my knowledge base couldn't find all 13 Jupyter tickets. Everytime I asked how many tickets contained the text "Jupyter" or "Jupiter", I got less than 13. I never got the full 13. I would get 2 of them, and if I modified the prompt, I might get 4. Usually, it was the same 2, but sometimes it was a different set of jupyter tickets. In frustration, I modified different settings including the Large Language Model, Source Chunks, and search type.

I constructed different prompt responses, including telling the LLM that it was reviewing Jira tickets and looking for similarities across the dataset. It didn't seem to matter what I changed, the results were all sub par, only giving me a small fraction of the responses I knew to exist.

If I modified certain settings, I got different results, and the best part was when I got responses with no mention of "jupyter" anywhere. The LLM so desperately wanted to give me a response that it gave me a nonsensical one.

On April 22, 2025, after a full year of proving I couldn't get back the 13 Jupyter tickets using the Opensearch Serverless Collection option, I tried a new tactic with the newly released Bedrock Knowledge Base using the Aurora Postgresql option. My use case started working! Mind Blown! What had changed?

Spoiler Alert - Two things changed:

My data changed. I no longer had access to my original dataset, so I created a synthetic dataset using ChatGPT.
I used the shiny new Aurora Postgresql option.

The functionality I had expected all along finally displayed on the page. The first question I asked to identify all Jupyter tickets returned a total of 13 tickets, and they were all unique, and actually matched the data from my CSV file.

Once I had proven I could get back those exact results, I started asking different questions and the results were significantly better. But was this just because I changed to a sql like database instance for a knowledge base?

No!

I created a new vector store database using opensearch serverless, and input the same data. I asked it the same questions, and surprisingly got back the full 13 jupyter tickets.

All things being equal, the only thing different was my data. When I reviewed the synthetic data I had created, I realized the description fields were significantly shorter than what existed in my original dataset. It would appear that the shorter descriptions made it easier to locate the specific tickets I was looking for. I then re-created the synthetic data to include longer paragraphs in the description field. I tested this out, and for the Aurora Knowledge Base option, it successfully returned the full thirteen tickets.

Summary

Amazon Bedrock Knowledge Bases are getting better at allowing me to chat with my data, but only when it's synthetic data. The Amazon Aurora Postgresql Knowledge Base option is more cost effective. The opensearch serverless option was costing me ~$5/day whether I was testing it or not. At this point, the functionality appears to be the same with my synthetic data across Aurora and the opensearch serverless option. Further testing is needed on a more realistic dataset to fully vet the capabilities of identifying trends in thousands of jira tickets.

Dante's Mirror - A Quantum Leap For Perception

AMELIA HOUGH-ROSS — Tue, 14 Jan 2025 02:23:49 +0000

How the AWS Game Builder Challenge, 2023 Valorant Championship, and Poverty Simulation Inspired a Video Game

How do you teach those with privilege to have sympathy for others?

How do you help people recognize when their behavior is in poor taste and is hurtful?

How do you support those open to having tough conversations that lead to better outcomes?

What if we try a video game? Not just any video game, one that everyone wants to play because it's always changing, and you never know what character you’ll be playing until you “win” the game.

When you consider the shift in perspective from the L. Frank Baum Wizard Of Oz story to the Gregory Maguire Wicked story, you gain a sense of how our perspectives can be broadened and understand there is more to most people's stories. For example, my social media feed recently shared a video of a woman in a sports car, and she is buying food using food stamps. The woman is ridiculed at the counter for driving the sports car and using food stamps. The story ends with the woman returning the sports car to a friend, because she doesn’t own a car, and only borrowed the vehicle to buy groceries for her family as she tries to get back on her feet. The judgments and ridicule experienced by the woman are seemingly unfounded now that we know the full story.

To help our society default to more optimistic perceptions, we need ways to make these stories mainstream. Additionally, if people only feel safe learning and exploring their light and dark sides when they are not being judged personally, then providing a video game is one way to explore this outlet.

AWS Game Builder Challenge

The January 14, 2025 due date for the AWS Game Builder Challenge created a sense of urgency around an idea I have been pondering for a few years. How do you get people to recognize they need to change their perceptions? Many people change when they’ve experienced a life-altering event, like the death of a loved one, escaping war, or some major incident that forces a perception shift. Most people don’t experience such heavy circumstances and can’t always recognize their privilege. How can we still provide an opportunity for perceptions to shift?

Where did this idea come from?

Attending the Valorant Championship activation in 2023 as an AWS Community Builder started as an exploration into eSports and was an incredibly educational experience. It also wound up triggering an impactful memory from 2018 when I participated in a life-changing poverty simulation thanks to the Missouri Community Action Network. This simulation is run with countless volunteers over a few hours and does an amazing job of putting the participant in the shoes of easily recognized scenarios of families down on their luck, or who have struggled to maintain savings and are just living paycheck to paycheck. I can't recommend this simulation enough! If you ever have an opportunity to participate, do it! (While the simulation is not meant to become a video game, ever since I participated in the simulation and noticed the heavy lift it is to run in person in real-time, the thought has been with me, that this could be a very impactful video game.)

The simulation provides an overview of about 10 different family units in which you participate where tough choices need to be made: Does your 15-year-old go to school today or work a job so you can buy groceries? It is based on real-world experiences and is eye-opening. The simulation immerses you in an alternate reality where tough choices, crime, and bad luck are the norm. This experience had a lasting effect and lingered in the back of my mind as my perception of people struggling and being unemployed changed. This simulation came to mind while visiting the Valorant Championship game. I wasn't at the one where Elon Musk helicoptered in with his son, but just that sentence alone is enough to highlight the fame, interest, and following that simply watching others play video games has.

The original plan to attend Valorant had to do with understanding the game itself, the art of the characters, and how they were dressed, as well as what impact this could have on gender and identity. I was completely shocked when I realized the enormous popularity of watching a team of people play a video game on the big screen. Dating myself, I think back to the film "The Wizard" with Fred Savage as he supports his younger brother as a video game-playing genius.

Valorant's huge fan following, and their interest in merchandise from t-shirts to stuffed characters had me thinking about how to channel this energy into changing people's perceptions. To inspire them to want to participate in a simulation to learn about those less privileged? What would it take to make a video game that could teach people what it feels like to walk in someone else's shoes for a day and hopefully learn something and gain insight to shift perceptions? This is the type of thing that keeps me up at night.

A video game is not the only idea or the best idea, but the AWS Hackathon presented itself and is an outlet for discussing this type of effort. In a world where bullying has been made easier thanks to technology, and when we no longer feel safe talking to our neighbors about bothersome behaviors, it seems we need to leverage technology to bring about some change. Additionally, we all could learn some new skills on how to better handle racism, sexism, and ageism in the home and workplace.

A year passed as I mulled over this idea, and I spoke with trusted friends who mentioned empathy games. My research timeline has been short, and while I’m familiar with the term empathy game, and have searched through STEAM to find a few games like “Passports Please”, “That Dragon Cancer” and “Papa & Yo”, I still have more work to do. Initially what struck me about these three games is that they are focused on severe situations and heavy topics like death, surviving war, and fighting diseases. These are not games you would play to gain skills in handling micro-aggressions. Perhaps, my game idea could help with day-to-day aggressions everyone is trying to manage. This hit home when my spouse was racially profiled in the church where our son attends daycare. The combination of this event and the AWS Game Builder Challenge put this effort in motion.

Micro-aggressions were not something I perceived as impacting me, but 2024 was the year that changed my perception. Working in technology, 2024 was the toughest year for me and I wound up experiencing multiple micro-aggressions that I didn’t fully know how to handle. People don’t seem to be aware of how their actions are perceived and how hurtful they can be, and my response after a few years of pondering is to pilot this video game idea based shamelessly on the Quantum Leap series, the poverty simulation, and the energy I saw around Valorant.

Basic Game Premise

Growing up and watching TV during the 1990’s, I enjoyed Quantum Leap! If you’re not familiar, the main character leaps through time, and embodies different people, but usually has no idea who he inhabits until presented with a reflection. The idea for Dante’s Mirror is like Quantum Leap, except you must search out the mirror. It’s hard to find and only presented after you’ve earned sufficient points to continue in the game. Points are awarded for handling tough situations and there isn’t always a clear direction. You only know what you look like if you can find a mirror and the game is structured such that you pass trials to get closer to the mirror’s location.

Game Mechanics

The game begins with the player in an unknown character in an unknown environment. There are no identifying characteristics on the player’s attire, and no way to identify yourself. You can only interact with your surroundings and respond when people talk to you or about you.

Example Scenario

You jump into a character working on a laptop in an open vestibule. People walk in and notice you, but don’t interact with you. They start talking about how “anyone can walk into this place”, and “should there be more security”, and a conversation ensues about the cost of security as these folks walk past you.
You look around and notice there is no one else in the room and you are alone.
What do you do?
How do you respond?
The game mechanics are such that you gain and lose points depending on the scenario and how you handle it. After navigating the room and finding clues, and interacting with surroundings, you gain enough points to find a mirror to look at yourself and discover you’re a certain gender/ ethnicity/ ability/ etc.

Next Steps

I found out about this game builder challenge too late to finalize a working prototype. Heading into 2025, I welcome feedback on the idea and would love to work with anyone interested in creating a pilot.

AWS All Builders Welcome Grant & re:Inforce Newbie

AMELIA HOUGH-ROSS — Mon, 01 Jul 2024 20:23:22 +0000

Earlier this year, I was excited to learn that I was one of ~fifty applicants selected to receive the inaugural AWS All Builders Welcome Grant to attend the 2024 AWS re:Inforce Conference held in Philadelphia, PA in June of 2024.

Once upon a time, I imagined my tech career would revolve around computer graphics. In reality, I have been slowly migrating to securing cloud applications. My latest contract work to support the future of CISA Secure Cloud Business Applications (SCuBA) earned me the opportunity to apply for the All Builders Welcome Grant. As a veteran attendee of AWS re:Invent, I found the re:Inforce conference accessible with more options to hop between sessions if my first choice was full. Join me in a re-cap of re:Inforce!

Monday, June 10, 2024

Arriving at the Conference on Monday to meet fellow All Builders Welcome (ABW) Grant recipients was awesome! This was not your typical meet and greet, and the facilitation was excellent. We divided up into groups of 10 so we could get to know each other and answered prompts about who we were, and what brought us to the technical security field. I made new friends and had folks to run into and eat lunch with for the rest of the conference. It was a terrific start to the day. During the break, before the conference started, I visited the Franklin Institute to check out the "Art of the Brick" exhibit.
If you love legos like I do, then it was a nice opportunity to see artistic expression using legos. I recommend the experience.

The remainder of the afternoon was visiting sessions, including:
COM322 - Improving your Amazon S3 Security with cost-effective practices by Brian Tarbox. This was a nice take on Infrequent Access and how it can be used for cost savings

TDR302 - Detecting and responding to threats in generative AI workloads. This breakout session provided an overview of securing Amazon Bedrock using GuardDuty and Security Hub.

APS303 - Strengthen open source software supply chain security: Log4Shell to xz. If it weren't for my colleague, I wouldn't have attended this presentation and it was definitely one of my favorites! I learned that Rust is the open source language with the most growth. I wound up meeting David Nalley after the event as part of the AWS Community Builders group and it turns out we live in the same state and I will be inviting him down for one of our Columbia AWS User Group meetings.

Tuesday, June 11, 2024

Today - I joined fellow ABW folks for breakfast at our hotel

and then went to the Keynote.

There were a number of interesting facts shared, specifically that S3 has over 350 Trillion objects. This was trivia in a later session and "Encrypting Everything" that I attended and confirmed people did remember what was said during the keynote.. I appreciated the guest speaker, Ash Edmondson from Eli Lilly. She is the AVP of Security Architecture & Engineering. I enjoyed her talk about using security cameras to check on her pet while she was traveling. It was a great analogy for why cloud security is important as well as the concept of Trust and the different levels associated with trusting your friend to watch your dog and the level of trust you want in a cloud service provider.

I also attended a great session, ABW101: Diversity & Accessibility in the age of Generative AI. It did an impressive job of highlighting how those with different abilities need to be considered in the GenAI discussion. Kudos to Amani Johnson for presenting a great talk!

I attended DAP331 - Encrypt Everything: How different AWS services help you protect data. This was a great session that reminded me why I studied so hard for the AWS Security Specialty certification. Most of what was discussed was from my studies for that exam, and made me feel like I needed to hurry up and recertify.

I joined the gathering of WiCyS folks in the Expo Hall and learned about how I could support them as a WiCyS member at an upcoming conference.

Additionally, I have been working on a use case with Generative AI, and found the "Chaos Kitty: Security Battleground Edition", BLD206, exhibit that used Amazon Bedrock to parse through security procedures to find poorly secured infrastructure. It was great to see this type of gamified learning enabled through Amazon Bedrock.

I ended the day in an Executive Security simulation for Responsible Generative AI Usage, TNC294. The team I joined came in last, which was great because we learned a lot about how AWS views the approach to Responsible AI usage. I am still waiting on the read out from the session, and I can't wait for it to be ready.

Wednesday, June 12, 2024

I wasn't able to attend many sessions due to work meetings that required travel to DC, but I am watching a few things on demand now that my travels have wound down.

Overall, this was a great experience, and a refreshing one to engage with more of the secure cloud technologies. The presentations I witnessed helped me prepare for my AWS Community Lounge speaking engagement at the AWS DC Summit. (More on that in another post)

Comprehending JIRA Tickets with Amazon Bedrock

AMELIA HOUGH-ROSS — Thu, 07 Mar 2024 22:27:05 +0000

Back in August, I wanted to gain insights from JIRA ticket data to better understand what our customers needed from our team, and if there were any trends I could identify to provide better self-service and FAQ documentation. I started that journey using AWS Comprehend which you can read about here: https://community.aws/posts/comprehend-jira-tickets

For those keeping track, Amazon Bedrock became generally available in September of 2023. My team had access to a preview, so when the AWS Comprehend entity analysis did not lend itself well to my use case; and I didn't feel like training a model, I started to get familiar with Bedrock. The following post is a follow-on to the Community article above and fleshes out a few details that will help those newer to Amazon Bedrock navigate the product.

First, getting into Amazon Bedrock from the AWS Console is pretty simple. You select "Bedrock" from the console and select "get started" and it takes you to a great Overview page where you can explore several foundation models. These foundation models are pre-trained by industry so you don't have to pay to train a model of your own. These models are ready to be applied to your use case.

To get access to a model and use it, that can be a bit counter-intuitive. You need to click on the "model access" link to see which models your account has access to, if any,

As you can see in the image below, some models are "available to request" and some others list "access granted". If this is your first time using Bedrock, most likely, you will need to request access to the models you are interested in for your project.

To finalize the request for the Large Language Model (LLM) of your choice, you have to click on "Manage model access".

Once you've clicked the checkbox next to the model you're interested in, you need to "save changes" and wait a few minutes for the model to become available in your account. AWS points out that you don't incur fees for using these LLM's until you've requested them and started using them.

Now that you have access to a specific LLM, you can start working with it.

For my Jira ticket use case, I leveraged the Bedrock Text Playground and selected Claude V2 as the LLM that I wanted to test using a sample of Jira ticket data. Leveraging the playground, I could drop a large amount of text, and then use prompt engineering to see what the Claude V2 model could pull out. I was happy to see that Bedrock's Anthropic option worked out of the box and appeared to support my use case in a way AWS Comprehend could not.

What continues to surprise me is prompt engineering. How you ask a question can drastically change the results returned by the LLM.

For example, if I ask Claude V2 the following question:
"Provide a list of tickets that contain text AWS"
Claude V2 replies with: "Here is the list of tickets that contain the text "AWS":"

BUT,

If I ask Claude V2 to "Provide a list of tickets that reference "AWS""
It responds with this: "Here are some of the tickets in the provided summary that reference AWS:"

The two ticket lists will include different tickets. Using Claude V2 for this type of analysis seems to require defining a vocabulary for questions that will elicit the response my human brain is expecting.

This outcome has been educational and highlights the opportunity to dive deeper into prompt engineering to ensure the Claude v2 output is what my human brain expects.

All this being said, it is eye-opening how crucial someone's understanding of a native language is to be able to determine if the results you receive from an LLM are truly accurate. You need to fully understand the data you're using and diagnose many different outcomes before you achieve the result you're looking for.

Stay tuned for my next writeup that will discuss how to work with Amazon Bedrock from VisualStudio Code.

Recapping Re:Invent 2023

AMELIA HOUGH-ROSS — Fri, 08 Dec 2023 01:37:17 +0000

The yearly AWS Re:Invent conference is truly a labor of love. Those who repeatedly attend look forward to it for many months, and then spend a few months recovering. After the American Thanksgiving holiday, we all descend upon Las Vegas and are quickly reminded of all the walking, networking, amazing content, and wonderful surprises that arise during the week.

Some of us old timers remember the year it was held in October and when it was small enough to fit into the Venetian and Sands Expo Center. This year, Re:Invent was a true joy because earlier in 2023, I joined the AWS Community Builders. In past years, I have usually been on my own, the only attendee from my organization, and I have spent much of the week making friends as I go, and eating with random strangers who inevitably become solid LinkedIn contacts. But it was always an empty feeling when I went home, never sure if the connections would materialize or continue into the next ReInvent.

The AWS Community Builders program has significantly changed my return from Re:Invent to one of invigoration. I felt fulfilled knowing my connections would lead to more projects and conversations.

To understand my journey this year, I’ve shared my daily play-by-play for Re:Invent.

Sunday:
Flying American Airlines was so painful this year due to mechanical issues on both my flights, missed connecting flights, and lack of access to food. After being awake for 12 hours, I finally made it to Las Vegas. This trip is normally a 5-6 hour direct flight. It didn't need to be this hard. Note to self, fly direct, to avoid connections and missing all the fun with AWS Community Builders.

Re:Invent CheckIn

Grabbing my badge at the airport was the fastest I can remember in recent years, luggage was waiting for me, and there was no wait for a cab, and the hotel was amazing enough to grant me a room early. I was wrecked after being up for so long and figuring out new flights, so I had a laydown and rested before going to pick up swag.

Swag Pickup

The peace of Sunday night in the Sands Hall was appreciated because hoodie and water bottle pickup were easy and uncrowded. I met up with a friend and noted Jeff Barr was meeting with people. I met Jeff for the first time this year and thanked him for his support of the AWS Community Builders. This was my first time riding the Datadog slide, with a picture worthy of cringe. Met with folks I only see once a year, and planned my schedule for Monday.

Monday:
There were so many choices for wonderful sessions all at the same time, that tough choices were made, however, the universe stepped in to validate my choices in many ways. First, I tried yoga at 6:30 am, which was one of the most popular events all week. Sad to say, the doors were closed when I arrived. Many other folks missed out on this as well. AWS got the message, and I think next year there will be a larger space to support mindfulness and a retreat space. Missing Yoga meant I was in the right place at the right time to talk to AWS contacts who recommended Experience Based Acceleration (EBA’s) as a means of migrating some on-prem projects into the cloud for sharing as open source. This looks like something my team will benefit from and I’m going to be exploring this in the coming weeks. Additionally, I have some new ideas for hackathons and ways to showcase our cloud-smart focus with the support of our leadership team.

My first session after breakfast was a workshop and thus not recorded. ENU301 – Using Generative AI on AWS to Accelerate Energy Solution Development. The session was not what I had in mind, not in a bad way, just different. It was more of an overview to understand what type and style of questions would lead to better results from Anthropic's Claude model. The session consisted of 2 parts and I was able to head over to Mandalay Bay for the Women in Tech programming.

Arriving at Mandalay Bay was super easy in the morning, and took ~15 minutes. Traffic was light. It took me a few tries to get my bearings once inside the facility since there is a "north" and "south" area, but it soon became easy to navigate and I attended a Women in Tech panel.
After the panel, many of us waited for an awesome networking lunch.

I met many strong women from all over the USA and we were discussing new employer benefits including egg freezing and reproductive support. These are great conversations and make the women's content so valuable. (At what other tech conference can you talk about reproductive benefits?) After lunch, I checked out the AWS Escape Room for Certified Cloud Practitioners and successfully escaped the toothbrush factory, and then spent some time in the Expo to chat with a few vendors.

Tuesday:
I chose to hang in my room to stream Adam Selipsky's Keynote so I could listen and submit a few month-end reports for work. I was happy that some of Adam's announcements were related to previews granted to AWS Community Builders. Check out the top announcements here

After the keynote, I traveled to MGM Grand for a luncheon with Pluralsight at the SkyLofts to visit with old friends, Faye Ellis and Drew Firment. I saw Ryan K. too! As a two-time Guru Of The Week from the A Cloud.Guru days, I enjoy learning what's changing within the Pluralsight platform.

)

On my way back from MGM, I saw Brooke Jamieson in passing and then ran back to talk to her. I shared my partyrock apps with her and she gifted me a friendship bracelet (secret swag) based on PartyRock apps created in advance of the Re:Invent launch.

In the evening I met with fellow AWS Community Builders at the Mixer in Buddy V's. This was a wonderful opportunity to meet new friends, say Hi! to old friends, and learn about future support within the Community program. Oddly enough, I received outreach to help set up a local User Group in my city. So all the things came together this evening for a wonderful night catching up over good food and drinks. There was even more secret swag, if you know, you know!

Wednesday:
Today was tough! I had a game day and a women's mixer with the Community Builders at the same time. The mixer won out as did playing the Builders Card Game and a return to Mandalay Bay for the first AWS Jam & Escape Room. During this Re:Invent, I was investigating gamified learning options to bring back to my organization.

Thursday:
I chose to listen to Werner Vogel's keynote, and found it enjoyable. I appreciated his sentiments and felt his message echoed discussions I've been having with peers related to cloud architecture, technical debt, and billing. Check out highlights here

After the keynote, I was scheduled to record a video for AWS Skillbuilder marketing which was a lot of fun to shoot; keeping my fingers crossed it will turn out well!

After the interview, I was back on the bus to Mandalay Bay for my final AWS Gameday with PartyRock. This was one of the better gamedays for me personally as well as for my team. We finished in the top 3 and had first place for a bit. I took on a challenge and was able to answer the questions, despite the coding portion not working out as I had planned. The whole experience was lots of fun, and I can't wait for next year's Re:Invent to see what new challenges exist. Jeff Barr even stopped by our GameDay!

Re:Play Party

Everyone knows Thursday night is the AWS Re:Play party with great performances and bigger-than-life games. The games were paired back this year, while the music was off the hook. I enjoyed dancing with friends to Major Laser and running into old friends from Cleveland. It was a fun night and one that didn't go too late, which helped me get to sessions on Friday.

Friday:
I usually depart on Friday to head home, but this year the flights were better to leave first thing on Saturday. This meant I could attend sessions on Friday, and I am glad I did! Hearing from
Clare Liguori was amazing! DOP228
Her story and presentation were engaging and I learned a few things about Q that I wouldn't have gathered otherwise. The session also allowed me to witness the excitement of many developers to hear from a woman in tech. This session and the experience, made my day!

I returned home on Saturday with minimal flight issues, although I came close to missing my connection due to delays. Again, note to self, fly direct next year!

AWS Community Builders Thankful Post 2023

AMELIA HOUGH-ROSS — Thu, 07 Dec 2023 15:48:31 +0000

Fresh off AWS Re:Invent, I am so thankful for the AWS Community Builders program and the Community Team! This program has made a huge difference regarding my connections as a "dancing engineer" working from home. Our "longest thread" friends (Johannes Koch & Jenn Bergstrom) from the Re:Invent hike shared the idea for Serverless Art and this SECRET initiative is ready for prime time with this post :-)

Thanks to Jenn for sharing her post on creating art using step functions, and for helping me play with something different.

PartyRock With Leftovers!

AMELIA HOUGH-ROSS — Sun, 19 Nov 2023 17:53:54 +0000

As an early Beta tester for the PartyRock , an Amazon Bedrock Playground, "Everyone can build AI apps with PartyRock, an Amazon Bedrock Playground"

If you are like me, you've probably been playing around with natural language processing and generative AI. The opportunity to preview the PartyRock Playground was a lot of fun, because it made generating an app built on top of a large language model, incredibly easy.

After a long day of work, you need to make dinner for the family, if you don’t have something ready-made, you search through ingredients and try to come up with a recipe.

When you’re all out of ideas, leverage #partyrockplayground to build an app to generate recipes for you.

This took me less than 3 minutes once I had the idea to feed ingredients and ask for a recipe. Sure, there are websites that can do this for you. The websites I've used in the past provided pretty involved recipes, and I was too tired for that level of effort. PartyRock was so easy to use, I had simple recipes in under 5 minutes.

For my first app, I was just so tired of trying to figure out what to make for dinner, and the ready-made food started to taste like cardboard after a bit. I usually don't have all the ingredients to support my existing recipe books, so with just a few ingredients, my "make a meal" app gave me yummy ideas for food.

I have many plans for my next round of apps including a kids toy recommendation based on age. (Just in time for the holiday shopping season).

I was thrilled to be invited to participate in the Beta, and as soon as the PartyRock Playground was launched, I had my recipe app running in under 5 minutes. I found the interface very intuitive, the longest time I spent was figuring out which app to create first. I am eagerly awaiting future updates and features!

AWS, Azure, and Google share spotlight on Cloud Day 2023

AMELIA HOUGH-ROSS — Tue, 01 Aug 2023 18:19:40 +0000

The Pacific Northwest National Laboratory (PNNL) Cloud Day, hosted in Richland, WA, USA, showcased an impressive array of technological innovations and real-world applications of #aws #azure #gcp.

PNNL is actively engaged in expediting research through the use of cloud service providers: AWS, Microsoft, and Google. In order to share what’s possible at PNNL, the laboratory hosts a yearly cloud day which took place in Richland, WA, this July 19, 2023.

The Cloud Day keynote from Kelly Marquardt with NEXTSilicon is a good reminder that hardware and software design for high -performance computing is key to unlocking the future of high -speed computing. I remember my own computer science training when designing software to work well on hardware was an important part of being an IT Engineer. The growth of computing during the past 20 years has led to this approach being less of a focus. When you need to do high -performance computing with exaflops though, the software / hardware integration can be the difference between ho hum compute and revolutionizing computation for researchers.

Cloud Day hosted many vendors including #SPLUNK #Atlassian #Gitlab and #IBM. Each cloud service provider had a day long track to share workshops, new product releases, and community learning sessions. The Google EarthWall made an appearance which is a nine monitor system that displays geospatial data supporting research efforts. Everything from agricultural impact of the war in Ukraine to rainfall impacts on flooding in the USA. Azure presented on Generative AI integrations and AWS shared a workshop using IOT Core (https://aws.amazon.com/iot-core/) to provide sensor data to a dashboard that can determine wind turbine effectiveness. Check out the workshop details here:

EnergyKit Workshop Public Link: https://catalog.workshops.aws/energykit-workshop/
EnergyKit Workshop repository: https://github.com/aws-samples/aws-energy-kit-workshop

Community interest groups named “Birds of a Feather” (BOF) took place in an open “hybrid” circle with a laptop and speaker in the middle showcasing how well PNNL includes its remote attendees. A great Serverless BOF took place where open discussions on Azure Functions, and AWS Lambda was discussed and debated based on use cases. A good time was had by all!

It is a terrific event and one that brought in over 900 people from across the USA as PNNL brought in a cadre of the remote workforce, vendors, and cloud service provider subject matter experts. This is a single day event that is part of a larger week -long #Techfest held yearly at PNNL.

PNNL is a research institution funded by the Department of Energy specializing in areas such as national security, nuclear materials, non-proliferation, nuclear fuel life cycle, and energy production.

keprtv.com