Streak’s AI roadmap

Artificial intelligence (AI) has become a buzzword in the tech industry, and for good reason. With its ability to learn and adapt to complex tasks, AI has the potential to revolutionize the way businesses operate. 

One area where AI has shown great promise is in software-as-a-service (SaaS) products. By leveraging the power of AI, SaaS companies can improve their products, make them more intelligent and efficient, and create better user experiences.

This is why AI will be a key component of Streak’s product strategy going forward. 

<div class="anchor-wrapper"><div id="why-ai-in-streak" class="anchor-target"></div></div>

Why we’re building AI into Streak

We’ve already started building AI into our product and many of our upcoming features will incorporate AI technology. This will help our users gain more intelligent and personalized insights, automate repetitive tasks, and learn how to use Streak features more quickly.

Since Streak is built on top of Gmail, we’re in a unique position to build really powerful AI features into a tool that you’re already using. Because our users trust us to securely add their emails to Streak, we can train our AI to do things like summarize large deals, find next steps, and automatically fill in data points about your deals. 

We have a nuanced viewpoint on the benefits of AI in our product - we think the primary benefit will be that our customers will have an easier time onboarding and setting up the product for their needs because they won’t need to be a Streak expert. We’ll be able to gather your intent and set up the perfect pipeline fully filled out with your actual data. We anticipate that our AI will get you 90% of the way there, and you’ll be able to customize the results to reach 100% accuracy. 

Another benefit is that we’ll be able to get far more data into your CRM automatically as you work in Gmail. From picking out opportunities directly from your team's inboxes to extracting data points (like deal size, anticipated close date, etc) automatically, your pipelines will be more accurate and actionable. 

<div class="anchor-wrapper"><div id="ai-features" class="anchor-target"></div></div>

AI features in Streak

We’ve already begun building AI into Streak. Below, we’ll share the features we’ve launched and are currently working on, as well as a glimpse at our AI roadmap for Streak.

<div class="anchor-wrapper"><div id="formula-generator" class="anchor-target"></div></div>

🤖 Formula generator (status: 💫 live)

The first AI feature we’ve incorporated into our product is the formula generator, which allows you to create custom logic and data points inside Streak formula columns without coding. 

Historically, we’ve supported adding small snippets of code to customize pipelines with data points specific to your business. For example, you could write a formula to calculate sales commission or return certain values based on data in your pipelines. 

Now, you can type a prompt in natural language and Streak AI will automatically generate the code and test it. Since it shows you what the code should be, it simultaneously teaches you how to use the product as well.

<div class="anchor-wrapper"><div id="deal-summarization" class="anchor-target"></div></div>

📝 Deal summarization (status: 🧪in beta)

Our AI deal summarization helps managers quickly review the state of a sales rep’s deal and identify decision makers and next steps. 

Instead of catching up on a deal by reviewing all the notes, emails and meeting history, in a Streak box, our AI summarizes and synthesizes all the communications. If you need more context, it can link back to the original content for more details.

This saves time for sales reps and allows managers to better understand what their direct reports are working on.

<div class="anchor-wrapper"><div id="semantic-search" class="anchor-target"></div></div>

🔍 Semantic search (status: 🧪in beta)

Semantic search will allow you to ask specific questions of an individual deal or set of opportunities. A sales manager can inspect one of their sales rep’s deals and ask questions like: “When did the customer say they could sign the contract?” or “How many seats are included in the contract?” 

These data points don’t have to be logged by the sales rep, instead the Streak AI will search through all the communication history and any meeting notes to find the answer.

<div class="anchor-wrapper"><div id="auto-filled-deals" class="anchor-target"></div></div>

🔢 Auto-filled deals (status: 🧪in beta)

Being inside Gmail is great because you can update CRM fields in the Streak sidebar while reading an email. However, it’s still manual data entry. 

With AI, we can extract data about a deal (like deal size, products, etc) and automatically suggest updates and tasks from the email thread. This turns data entry into just a single-click confirmation.

<div class="anchor-wrapper"><div id="intelligent-import" class="anchor-target"></div></div>

📱 AI-generated call and meeting agendas (status: 🧪in beta)

Create a focused agenda for your next call. Streak can create a call or meeting agenda based on your past conversations and deal history to help you move the deal forward, or add a goal or topic for the conversation to create a custom agenda.

🙎Intelligent import (status: 🧪in beta)

When users first start with Streak they have no data in their pipelines. To get started you need to import data from another source, which involves lots of data wrangling, or manually enter the data. 

Instead, AI can scan your inbox and intelligently pick people and companies that correspond with your use case of sales, hiring, etc. This will differ from other tools, which just add everybody from your inbox indiscriminately.

<div class="anchor-wrapper"><div id="natural-language" class="anchor-target"></div></div>

🔪 Slice and dice with natural language (status: 🧠 R&D)

Streak already lets you slice and dice your pipeline to create filters and segments with our saved views feature. 

Currently, you can build saved views with a series of drop-down options to configure filters, sorting, and grouping within your pipelines. While this allows for endless segments and filters, it can be confusing for new Streak users.

With AI, you will be able to ask Streak to create the segment or filter in natural language: “Show me all deals created in the last 30 days that are in the negotiating stage where the customer is waiting on an email from us. Show the biggest deals first”.

Similar to the formula generator above, the AI will not only create the correct saved view for you, but it will show you what the inputs should be in the saved view editor so you can learn how to use the feature in the Streak UI as well.

<div class="anchor-wrapper"><div id="generate-email-sequences" class="anchor-target"></div></div>

✉️ Generate Email Sequences (status: 🧠 R&D)

Email sequences with automatic follow-up have been a game changer for Streak users who need to send mass emails and perform email outreach.

AI can make our mail merge feature both more efficient and powerful. During sign-up or when creating a mail merge, Streak can ask users basic questions about their business and generate high quality, highly converting emails sequences with optimized automatic follow-ups. 

By just answering a few questions, Streak can generate the specific customized content for each recipient as well as automatically follow up at the right cadence.

<div class="anchor-wrapper"><div id="draft-replies" class="anchor-target"></div></div>

✍️ Draft replies (status: 💡Idea)

When viewing emails from leads and opportunities, AI can automatically draft replies based on how you’ve responded in the past to similar deals. Of course, you’ll still have the opportunity to customize the message and confirm the details before hitting send.

<div class="anchor-wrapper"><div id="pipeline-autopilot" class="anchor-target"></div></div>

🧑‍✈️ Pipeline autopilot (status:💡Idea)

Defining a workflow and giving it structure helps users know what to do next, make sure nothing slips through the cracks, and ultimately convert more deals and opportunities in their pipelines.

We can use previous conversion history for your team to simplify a salesperson’s job to a series of yes/no questions. The AI would suggest data updates, which deals should be followed up on, draft replies to recent emails, and autofill data all in a single interface. 

This is the sum of previous AI projects with a UI that leaves the salesperson just directing the autopilot from a single point in the product.